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.
Similar content being viewed by others
Introduction
Migratory birds move seasonally between breeding and non-breeding sites in response to environmental changes and genetic cues [1]. In visiting multiple areas across seasons, they encounter various pressures including unpredictable environmental conditions and resource availability, hunting and predation risks, and anthropogenic degradation of migratory routes. Combined with the energetic challenges of such a long journey, these pressures make long-distance migration particularly risky [2]. In fact, it has been shown that both habitat loss during the breeding season, and anthropogenic and climatic factors during the non-breeding season, are the main causes of large-scale declines in long-distance Afro-Palaearctic migrants since the 1970s [2, 3]. In addition, lower survival has been described for some migratory birds during migratory periods ([4, 5]; but see [6]). Yet pressures happening during the breeding, migration, and non-breeding periods do not solely result in a direct increase in mortality. Events occurring at one stage of the migratory cycle are also known to influence the breeding success or survival of individuals in subsequent stages, a process known as seasonal interaction [7]. Thus, a clearer understanding of the ecology, spatial dynamics, and demography of migratory species requires deciphering the links between populations at different stages of the migration cycle. Such links, drawn by the periodic movements of migratory individuals between populations or areas, have been defined as migratory connectivity [8].
The importance of migratory connectivity for conservation biology has been emphasised from the very beginning, and has found support in a diversity of studies since then [8], becoming a cornerstone for international initiatives such as the Convention on the Conservation of Migratory Species of Wild Animals [9]. Beyond bringing the long-awaited data enabling the population dynamics of different seasons to be connected [10], migratory connectivity is potentially instrumental in the transmission of pathogens [11]. It can also help correlate localised environmental changes or anthropogenic pressures and distant population declines [12, 13], understand the robustness of migration networks (e.g. [14, 15]), and contribute to identifying conservation hotspots. Migratory connectivity influences how the spatial structuring of populations and the movement of individuals may affect species susceptibility to global changes [8, 16]. It is thus among the main priorities for research to devise successful conservation measures for migratory species [17,18,19].
Multiple methods have been used to identify the connections between the breeding and wintering grounds of migratory birds [8]. Bird banding has been used since the beginning of the twentieth century to track individuals between their marking and reencounter sites [20]. Stable isotopes and genetics have been used to study migratory connectivity more recently [21,22,23]. Such markers, referred to as “intrinsic markers”, are naturally carried by every individual and can provide information on the geographic origin of the individual without needing to mark it. In parallel, the development and miniaturisation of tracking devices such as radio and satellite transmitters, geolocators, and the establishment of large-scale telemetry networks such as MOTUS [24] and the Icarus initiative [25], have enabled explicit spatial data about migration tracks to be collected [26, 27].
The continuous improvement of these key approaches over the last 20 years has led to a significant increase in the number of publications on migratory connectivity (Fig. 1). The expanding toolbox of approaches to investigate migratory connectivity has allowed these studies to diversify and start investigating both fundamental questions about the establishment and consequences of connectivity patterns (e.g. [28]) and applied questions about species conservation and epidemiology (e.g. [29]). Additionally, the growing emphasis on data integration in ecology has fuelled this trend, with the potential to offer a major turning point in broadening our perception of migratory connectivity by allowing the construction of direct bridges between connectivity and population or environmental studies. Indeed, data integration appears to be a promising tool for bringing together what are currently considered to be separate fields of biology [30].
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
With the multiplication of studies investigating connectivity patterns along with en-route behaviour (e.g. [31]), environmental changes (e.g. [14]), or population dynamics (e.g. [15]), the field of migratory connectivity has never been as close to other ecological fields as it is now. Yet if promising applications have already highlighted the potential of connectivity for future research, much of the data remains to be collected. We believe that, for the field to fully contribute to answering broader ecological questions, transversal studies are needed, which may benefit from making the methodologies to investigate migratory connectivity more accessible. Here, we provide an overview of the latest advances to investigate migratory connectivity and discuss how these methodologies are developing. A particular focus is the promise, as well as the challenges, of data integration, a topic that will become increasingly important as migratory connectivity is applied in a wider range of contexts.
To establish this synthesis, we scanned through the literature returned by Google Scholar and the Web of Science using the broad search pattern “migrat* connectivity”, in combination with a series of keywords referring to the various types of data and data combinations used in the studies (see Additional file 1). We could then focus on (1) articles pointed out by several other publications as establishing major advances in each approach, in order to identify the most recent methodological developments; (2) major articles and reviews giving extensive information on each individual approach for further information; and (3) articles using a combination of data to investigate connectivity patterns, in order to build up a picture of the existing combinations and their use.
Tools to investigate migratory connectivity
Bird banding
Traditionally, birds captured on their breeding or non-breeding grounds have been marked with leg bands or other markers, indicating their place of origin if they are later recaptured, resighted, or their corpse is recovered elsewhere [20]. While banding data has long been used to draw qualitative connectivity patterns, it is now the main data allowing quantitative links between populations to be estimated (Box 1).
This shift from qualitative patterns to quantified estimates of migratory connectivity relies on statistical models capable of disentangling the probability of movement between populations (i.e. connectivity parameters) from survival probability and spatial variation in banding effort and reencounter probability, which are intertwined in raw data [32]. These probabilities serve to express the general probability of reencountering a marked bird in a given area, depending on the type of available data—live reencounter or dead recovery—in multi-state mark–reencounter models used to investigate the movement of individuals between predefined states (i.e. sites and/or seasons in this case) [32,33,34,35]. The division coefficient method offers a straightforward alternative to quantify migratory connectivity over long time periods by estimating movement probabilities between sites [36]. The aim of this method is first to estimate reencounter probabilities in each destination from simple systems of equation, assuming that these probabilities are the same whatever the origin of the individuals. Knowing the number of birds ringed in each source area, the number of birds reencountered in each destination and the reencounter probabilities then enables the calculation of the movement probabilities.
Recent methodological developments exploit the flexibility of mark–reencounter modelling frameworks to deal with the spatial and temporal heterogeneity of banding data, which stems from the unequal number of observers or observations between seasons and places (Table 1A; see [37, 38], for a review). These recent approaches can take into account sources of heterogeneity such as seasonal variation in reencounter probability [34] or can integrate new information known to affect reencounter probability such as human density in the recovery region [39]. Models have also even been improved to work when information on the total banding effort is lacking [35, 39, 40]. Overall, these developments have been carried out using Bayesian modeling, which is particularly flexible for building complex models that notably take various sources of uncertainty into account and allow for the combination of various types of banding data, favouring numeric estimations (e.g. [34, 35]).
Nevertheless, the number of reencounters generally remains very low—e.g. only two birds in Giunchi et al. [51] and barely 1,000 out of 500,000 birds tagged over 50 years in Korner-Nievergelt et al. [34]—and the spatio-temporal heterogeneity in observation effort tarnishes all available datasets [38]. Although these drawbacks can be partially offset, taken together with the complexity of the required modelling approaches, many studies remain qualitative (e.g. [52, 53]).
Tracking technologies
Tracking technologies include electronic tags such as transponders or spatial data loggers fitted onto captured birds. Archival geolocators (e.g. [27]), satellite telemetry (e.g. [26]), and radio telemetry (e.g. [54]) are the main tools used in migratory connectivity studies. They generally provide precise and immediate information on the whole migratory route, including breeding, wintering, and stopover sites, and are thus very powerful to draw precise yet qualitative connectivity patterns—although methods to quantify migratory connectivity have recently been adapted to include scarce tracking data (Box 1).
Light-level geolocators, which derive latitude and longitude from the relative duration of day and night and the occurrence of solar noon, are widely used due to their low cost and weight [29, 47]. However, their use is limited by their particularly coarse spatial resolution (up to 300 km), which can be partly offset by the use of standardised filtering methods to remove aberrant locations, although this poses other issues in terms of reproducibility [62]. Also, the data they provide suffers from a recovery bias, since such archival tags need to be retrieved to collect the data [27]. Recent developments have aimed to increase their resolution using modelling approaches [63, 64] or through technical improvements such as the addition of light stalks that better record light intensity [41] or the conception of alternative archival GPS tags [42]. Statistical models can also be used to account for recovery bias ([12]; see [43], for a review; Table 1B).
Progressive technical advances in tracking technologies have resulted in a reduction of the cost and weight of devices—especially for satellite tags with various transmission systems (Argos, GSM or radio)—allowing more individuals and smaller species to be tagged [25, 44]. The development of continental-scale receiver networks is also likely to promote radio telemetry as a major tool for investigating migration [24] (Table 1B). The use of movement and behavioural models enables to further improve the high spatio-temporal resolution of the tracking data and obtain insights into migratory connectivity down to the scale of migratory routes [65]. And beyond telemetry, the application of machine learning to infer migratory movements from weather radar data may even enable the tracking of migratory populations on a continental scale, especially if species can be discriminated in the data [66].
Yet the cost of tracking devices remains high, leading to problems in representativeness and statistical power due to the low number of tagged individuals. As a first solution, rigorous sampling designs should help reduce bias in data collection [67], for instance by sampling individuals of the whole breeding and/or non-breeding distribution range of a species in a statistically randomised sampling design [68]. This has even been suggested as a way to reduce bias in connectivity measures [69]. However, putting the same sampling effort into remote areas as into accessible areas may add to the cost of the method and requires international collaboration. Also, the sampling design will still be limited by capture biases (i.e. which individuals are more susceptible to be captured) and monitoring biases (i.e. which individuals can be tagged). The weight of tracking devices induces a bias in our perception of migratory patterns towards larger taxa or larger individuals within a species, and attention should be paid to a potential effect on bird survival and flight performance [70].
Genetics and genomics
Individuals using the same breeding grounds are more likely to mate and share similar genetic features. Thus, analysing the genetic diversity of samples collected on a migratory site may allow the probabilistic assignment of individuals to a reference breeding population [22] and the identification of broad connectivity patterns (Box 1).
Investigations of connectivity patterns based on genetic data have been conducted using population genetics tools on various types of DNA sequences, the most recent being mitochondrial DNA, short repeated sequences such as microsatellites, and Single-Nucleotide Polymorphism (SNPs; [8, 56, 71, 72]). All these sequences are capable of carrying information on the genetic similarity or differentiation of individuals to varying degrees. If collected on the non-breeding grounds, it is thus possible to know if two individuals come from the same breeding grounds. If reference individuals have been sampled directly on the breeding grounds and if the marker shows a geographic structure, genetic assignments can be used as geographic assignments. Therefore, an investigation of population differentiation based on discriminant analyses [29] or differentiation metrics such as the fixation index (FST) [56] can offer initial insights into the relationships between breeding and non-breeding populations. Yet most often, the breeding origin of migratory birds is determined using well-established clustering algorithms followed by one of various assignment methods (see [29, 73] for examples).
Today, technical advances enable the exploitation of genomic data [45, 73, 74]. Species-specific markers based on microsatellites or single-nucleotide polymorphisms (SNPs) can now be developed more easily [46, 72]. In this way, genomics opens the door to larger datasets capable of providing more information about the genetic structure of populations at ever-decreasing costs (Table 1C).
Nonetheless, genomic data should be handled with care, as not all parts of the genome are informative [45]. The reliable clustering of individuals requires some genetic differentiation between populations, which can be rare for animals with high dispersal ability [63]. In this case, genetic differentiation may concern only a few loci, information that can be diluted when too much irrelevant genomic data is added. Even for species having a strong genetic differentiation, identifying sequences that can be considered characteristic of a particular area or population requires extensive work to collect and analyse high-quality reference samples over the entire breeding range of the species. Additionally, despite declining costs, sequencing such a large number of samples remains costly and may pose some logistic issues due to the transport of the samples between countries. So, while genomic data has unquestionably advanced our ability to detect population differentiation and to make assignments, this approach still has inherent limitations.
Stable isotopes
Geochemical processes result in spatial variations in stable isotope proportions within natural environments [75]. For instance, the condensation rates of heavy and light isotopes are different, leading to a spatial variation in the proportion of 2H and 1H along the clouds journey. These processes are further modulated by local variations in temperature and elevation, as well as biological processes, which paint a complex landscape of varying isotope proportions [75, 76]. These proportions are recorded in tissues such as moulting feathers [21, 23] and growing claws [77] through the uptake of food, with again some bias due to differential partitioning of light and heavy isotopes during biogeochemical processes. Individuals with similar isotopic signatures in such tissues can then be assumed to have grown them in similar isotopic environments, and thus to come from the same geographical region. Mainly used to draw large-scale connectivity patterns, isotope data is in the process of being included in quantified connectivity estimates (Box 1).
Isotopic signatures can be used to assign individuals to populations of origin using clustering tools such as discriminant analyses [78]. Knowledge of the geographical distribution of isotope values, i.e. the isoscape, further allows individuals to be directly assigned to regions using a variety of methods. These methodologies generally rely on the construction of a calibration regression model to correlate isotope values in the environment and isotope values in biological tissues of known origin (see [79], for a review). In particular, the probabilistic assignment of individuals on continuous geographical surfaces of isotope values is becoming popular ([48]; see [80], for a review).
The strength of the isotope approach is that huge datasets can be easily collected and cheaply analysed to obtain wide-scale connectivity patterns for several species [81, 82]. δ2H is the most common isotope ratio used to study migratory connectivity due to its clear latitudinal patterns and well-characterised isoscape described as part of long-term hydrological surveys [21, 75]. Other isotopes such as δ13C, δ15N, or δ34S are increasingly used to get more accurate spatial assignments when δ2H patterns are too weak [49, 83, 84]. Additional research has focused on means to account for inter-individual variation in fractionation and large local variation in isotope values, both of which contribute to blurring assignments ([47, 76]; see Table 1D). Applied to museum collections, this approach allows historical assignments to be performed to study variations in migratory connectivity in the face of global change [85].
Despite these developments, the isotope approach is limited by low spatial resolution, as birds are assigned to areas often spanning thousands of square kilometres (e.g. [82]). This low resolution is further amplified as probabilistic assignments only partially take into account uncertainties in measurements and models [80].
Beyond these four main approaches to investigate migratory connectivity, other methodologies have occasionally been used to extract information on animal movements between migration stages (see [50] for a more comprehensive overview). These methodologies notably include assignments based on phenotypic variation between source populations (e.g. [86]), on trace elements yielding a higher resolution than common isotopes [87], or genetic analyses performed on parasites carried by migrating individuals (e.g. [40]). Recent modelling approaches even allow accurate movements between populations to be extracted from abundance and citizen science data in some specific situations [88,89,90]. Given the rise of citizen science, they offer great potential to push forward and/or complement migratory connectivity studies based on more traditional approaches.
Combining data in migratory connectivity studies
Combining different connectivity data emerged early in the 2000s as a means to buffer some of the drawbacks of each approach [49, 71, 91] (Fig. 1). A whole range of possible combinations have been used in recent studies, incorporating up to three of the previously described approaches along with complementary data (e.g. [92]) (Fig. 2.1). Combined analyses can be classified into two main families: those that pool information from different analyses performed separately, and more recent integrated approaches that analyse data in a single modelling framework.
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
Separate analyses of different data types
Combining information from separate data analyses was the first form of data combination to appear in migratory connectivity studies [71, 91]. The benefits resulting from such combinations depend on the degree to which the different data contains the same information (Fig. 2.2).
Most often, studies compare the consistency of results derived from different data types (e.g. [29, 52]). This can be a way to increase sample size, particularly for species for which very few banded birds have been reencountered or few tracking devices fitted [51]. The comparison of different data types can also prove useful when one data type is insufficient to detect connectivity patterns. In the case of golden-winged warblers (Vermivora chrysoptera) and blue-winged warblers (Vermivora cyanoptera), tracking data revealed a stronger connectivity pattern than that obtained from genetic analyses, highlighting the unexpected low differentiation of previously used genetic markers [63].
Some studies focus on the complementarity of the information resulting from different types of data rather than simple consistency in results. Each data type can provide information on different subsets of a population. As pointed out by Finch et al. [53], band recoveries provide information on juveniles and individuals that failed their migration whereas tracking data is biased towards successful and heavier older birds. Different types of data can also provide information at different scales and resolutions. The resolution of tracking data can, for instance, serve as a means to check the accuracy of isotopic assignments [47]. In the case of genetic and isotope data, the complementary information they respectively supply about longitudinal and latitudinal location improved the resolution of spatial assignments in North America [71, 93].
Precise information obtained with one data type has also been used as a constraint in the analysis of another data type, e.g. by restricting the area of interest for spatial assignments [83]. This is appropriate for increasing the signal-to-noise ratio by avoiding the inclusion of irrelevant regions. Isotopic spatial assignments notably benefit from tracking data [29, 83] or knowledge about the distribution range of the studied species [29, 48, 82], which can be used to limit the final assignment area, improving the spatial resolution of analyses.
Integrated modelling
Data integration sensu stricto is based on a model that inputs a variety of data that is analysed within a single framework. An integrated model can be seen as composed of multiple submodels—one for each type of data—which are combined in a core equation. If the different datasets contain independent information on a similar variable, then this variable appears in each of the corresponding submodels, and can thus be estimated from all these data at the same time. Schematically, the information from the various datasets is crossed, allowing for a better isolation (or “identification”) and estimation of that variable (Fig. 2.2). Although the foundation of integrated approaches in migratory connectivity studies was established in the first half of the 2000s [49], it really began to emerge in the last decade (Fig. 1). Most integrated connectivity analyses are conducted in Bayesian frameworks due to the inherent flexibility of the underlying algorithms (Box 2).
Integrated models were first used to include various types of prior information, which act as a smooth constraint on the analysis parameters based on prior knowledge of the studied system [49]. This method allowed new types of data to be combined with common banding, tracking, isotope and genetic approaches to further increase the accuracy of results [49]. Data on breeding abundance is, for instance, integrated as a prior since the probability of originating from a population does not only depend on the frequency of the marker of each population of origin, but also on the relative size of these populations (e.g. [49, 82]). Other complementary data include knowledge of migratory directions based on banding data [48, 94] or population genetic structure [95], as well as information on species distribution [74, 96].
Studies that quantify migratory connectivity (i.e. where connectivity takes the form of an estimable parameter such as transition probability between sites, see Box 1) can fully integrate different data if the statistical models used to analyse the data share common parameters [57] (Box 2). The Bayesian framework is unequivocally preferred since it allows the straightforward combination of probabilistic models (see, however, [97], for another statistical framework), provided datasets provide information on the same population and are independent [57, 98]. By extending the structure of banding multi-state models to simultaneously analyse tracking data, Korner-Nievergelt et al. [57] significantly increased the precision of their connectivity estimates. Integrated analyses are more powerful than independent analyses in disentangling the intersecting information in datasets—the shared parameter(s)—from the data-specific uncertainty, allowing more information to be extracted from each dataset. Integrated models have been used to jointly analyse banding and tracking data [12, 39, 57], as well as combinations of banding, isotope and parasite data [40] or isotope and genetic data [74, 95, 98].
Additionally, integrated models have shown interesting results in other fields of ecology, such as the ability of integrated population models to estimate uninformed parameters [99]. This property originates from the fact that some data bears information on various processes altogether—e.g. survival, fecundity, and migration rates for population counts. These processes can be isolated and identified by crossing the entangled mass of information of population counts with the more specific information brought by other independent sources of data (e.g. breeding monitoring, capture-recapture). If the specific information brought by the other data allows isolation of all but one of the processes, this last process logically appears isolated and identifiable too. It would be reasonable to think that population counts for migratory species also bear information about migratory connectivity, which is supported by recent work modelling migratory connectivity directly from abundance data [89, 90], in the same way as the various connectivity data bear information about demographic rates (e.g. survival for banding data). Integrating population models and migratory connectivity models would be an interesting lead to follow to push further our understanding of the interplay between population dynamics and migratory connectivity. Such properties—associated with the wide range of possible data that integrated models can jointly analyse—suggest that they are among the most promising tools for migratory connectivity studies.
Challenges of combining data in migratory connectivity studies
The diversity of available methods raises questions about which type of data and/or combination should be favoured, and how to optimise their use. The performance of different analyses is generally compared based on the precision of spatial assignments [100] or connectivity estimates [57]. The accuracy of results has also been assessed by evaluating the assignment of samples or individuals of known origin (e.g. [95, 97]) or by comparing the results to high-resolution tracking data (e.g. [47]).
What data to prioritise?
To our knowledge, no study has provided an explicit answer as to which data should be favoured in migratory connectivity studies. How to optimise the sampling design to get the best description of connectivity patterns needs to be addressed, as collecting data for more species to build connectivity atlases will be a costly and time-consuming process [82].
Since migration leads birds to occupy different places at different times, often covering large geographic areas, it may be worth questioning where and when data should be collected to draw the best picture of connectivity patterns. Some data may be more informative and have a greater signal-to-noise ratio if collected during a specific season. For example, geolocator data shows higher latitudinal accuracy at higher latitudes, so tagging individuals on non-breeding grounds at low latitudes and collecting data on their movements towards the breeding grounds at high latitudes should be more informative than tagging birds on the breeding grounds and tracking them to their non-breeding grounds [101]. Furthermore, quantified estimates of migratory connectivity are sensitive to sampling designs and can be biased if trying to make an inference at a broader extent than the sampled area [69]. Some studies have started providing guidelines to understand the effects of the sampling design on connectivity estimates and adapt it to the study system [102]. More generally, applying statistically randomised sampling designs across the migratory cycle may be a promising direction to improve estimation [68]. Most studies rely solely on data collected during either the breeding or non-breeding season, or autumn or spring migration (e.g. [39, 103]). Yet study robustness may benefit from sampling at breeding, non-breeding and stopover sites, which may help avoid sampling bias, especially when populations mix between seasons (low connectivity strength, see Box 1; [8]).
As for the type of data to collect, all approaches used to track birds balance the scale and resolution of data in different ways. Tracking data is increasingly used to get a picture of migratory connectivity at high resolution, even at the scale of the migration route [54], which also holds promise for behavioural studies [28, 31]. Yet the ability of isotope data to describe connectivity patterns for hundreds of individuals across several species [82] and even over long time periods [85] is incomparable regarding the scale and representativeness of results. The complementarity of these two data types illustrates why no one today advocates for the use of a single data type. This is further explained by the restrictions and advantages of the different methods to collect data, which depend on the characteristics of the species. For instance, tracking technologies are usually limited by the size and weight of the individuals, although the progress of telemetry, light loggers, and even GPS tags now allows tracking of passerines [104]. Isotope data is mainly usable for species that grow their tissues in areas with clear isotopic signatures or gradients, as it is the case for species breeding in North America [82]. Similarly, genetic data is highly informative if breeding populations have a strong genetic identity that can be characterised [63], which is more probable for species with low breeding dispersal. Finally, bird-banding can be used for every species as long as the detection and capture probabilities are not too low and that large-scale effort can be put into band reading.
The fact that, currently, no approach is clearly superior to another instead shifts the question to whether all data can be combined, and whether certain combinations are more effective than others. For instance, the high longitudinal resolution of geolocator data would pair well with the good latitudinal resolution of hydrogen isoscapes in North America or Europe [82, 101]. This question has become even more important through the years as the combination of connectivity and more diverse complementary data, such as spatial occurrence, have proven to significantly improve the scale and resolution of the migration patterns we can draw [105]. Consideration needs to be given to the constraints associated with differences in scale and the representativeness of the data to be combined. Despite being put forward as a strength of combining data, mismatches in scale and sampling bias need to be handled with care to avoid further skewing results [106]. It is all the more important to consider these questions as new data combinations may appear with the development of integrated approaches.
How to optimise the combined analysis of different data types?
Investigating the performance of independent versus combined analyses suggests that not all combinations are relevant. In their study, Ruegg et al. [74] showed that the precision obtained with genetic data alone was almost equivalent to that obtained when combining genetic and isotope data with predictions of habitat suitability models. More strikingly, Hobson et al. [82] and Reese et al. [107] showed that including abundance data can sometimes worsen assignments. Indeed, abundance data can be particularly spurious when populations occur in patches because models then tend to neglect information about individuals that come from low density regions [103].
A proper understanding of what data is worth combining, and how to do it, requires understanding how each data type can contribute to the final results. This contribution should not necessarily be equal due to differences in the quality and quantity of the data. By quantifying the overlap between the posterior distribution of independent models with the posterior distribution of the integrated model, Korner-Nievergelt et al. [57] and Von Rönn et al. [40] observed that the contribution of each data type to the estimation of a given connectivity parameter was related with the actual amount of data they had for this region. In other words, for each parameter, the integrated models relied more on the most abundant type of data that could inform on it. Korner-Nievergelt et al. [57] stressed the need to improve current models to account for sampling effort and thus release the strain exerted by the most abundant data, which is not necessarily more informative than a smaller dataset sampling rigorously the study system.
The first way to optimise combined analyses would be to investigate which combination method performs best given the data at hand. As shown for species distribution models, the combination method can indeed affect the accuracy of the results [108]. Different combination methods give different weights to the various data (see Fig. 2.2), so the choice of a method is a first way to account for differences in quality and quantity of the data [108]. Future studies would thus benefit from a general comparison of the different types of combination that have been used so far, as well as a more specific assessment of how well various combination methods perform for their study system.
Once a combination method has been chosen, it is possible to adjust it to further control the contribution of the data to the results. Various attempts have been made to weight contributions of multiple types of data within a model according to the confidence researchers had in the different datasets. For instance, Ruegg et al. [74] and Rundel et al. [98] raised the probability of origin obtained for each data type to different powers when combining them using Bayes’ rule, tuning the parameters to optimise the spatial assignments of individuals of known origin in a cross-validation procedure. However, as stated in these two studies, such weighting may differ between groups of birds and be dataset dependent. The added value of weighting data has been highlighted using optimisation methods. Rushing et al. [100] found that for some species, correct weighting of isotope and abundance data decreased the surface and error of assignment, but incorrect weighting could increase both of these, or cancel out an increase in precision by a major increase in error rate (or vice versa). Combining data therefore can give less accurate results than the analysis of a single data type if not carefully tested using known source samples. More work is thus needed to assess whether the contribution of each data type should be modified and how to do this objectively [106].
Which approaches to integrate data across the migratory cycle?
Migratory connectivity has important implications for population dynamics, and is crucial to predicting the impact of climate change on migratory birds [19] and devising conservation measures [17]. Indeed, migratory species are particularly sensitive to site degradation and to localised events that may affect the survival or reproductive success of individuals, either directly or through alterations to the body condition or phenology of individuals that carry over to the subsequent seasons [7, 109]. The consequences of such events that affect only part of the migratory populations cannot be understood without a proper knowledge of where individuals originate from and where they go. Investigating the relationships between migratory connectivity and demographic and environmental processes has thus become a natural direction for recent studies about migratory species. For instance, information on migratory connectivity has been combined with demographic and environmental data using demographic models to understand how connectivity affects population counts and demographic parameters [110]. Conversely, migratory connectivity has also been investigated in light of network theory to understand how changes in populations or sites might affect the whole migratory network [15, 111]. However, information about migratory connectivity is still lacking for the vast majority of species and the methods to include such data into population dynamics analyses are still developing.
Current demographic models generally consider migratory connectivity as a static input to estimate the degree to which changes in a given region may affect population dynamics rather than as a variable ecological process [110, 112]. To our knowledge, connectivity has only been thought of as dynamic in theoretical network models where it can be estimated as an equilibrium output affected by changes in the migratory network [15, 113]. Such dynamics need to be understood now that we are able to more clearly describe patterns of migratory connectivity. The integration of connectivity and population data is likely to bring insights on the effects of connectivity on population and migratory network dynamics—and potentially the inverse relationship. This data integration shows promise to further advance the performance of full annual-cycle population models that consider events in the breeding, non-breeding and migration periods, and their consequences on population dynamics [10, 114]. The development of such bridges between migratory connectivity, environmental, and population dynamics tools has great potential to improve our ability to conserve species.
Conclusion
Technical and methodological advances over recent decades have greatly improved the quality, reliability and accessibility of the bird movement data that underpins migratory connectivity patterns. In parallel, the diversification of methods to combine data has further boosted the precision and accuracy of patterns and estimates, as well as their spatio-temporal resolution and scale. The various methods for tracking birds through space, and for combining this data with complementary information, are two important dimensions to be considered in the toolbox to investigate migratory connectivity. Data combinations have not been systematically used in previous studies mainly because only one dataset had been collected for the species of interest, or because only one type of data was informative enough for the study. Now that large-scale data collection must be carried out to build the long-awaited connectivity atlases [82], the potential of combined analyses should be kept in mind. However, the diversity of methodologies that now compose this toolbox pose a number of questions. No approach currently seems to outperform the others, making it hard to navigate through the jungle of methodologies. In this context, further work is needed to assess the performance of these methodologies, to determine where and when connectivity data should be collected and how to get the best of data combinations. Answering these questions is even more important as migratory connectivity starts to be considered in a broader range of ecological questions. New data are susceptible to be analysed along with migratory connectivity to further understand the behaviour of migratory species, their population dynamics, evolutionary history, and sensitivity to global changes. Notably, data integration has the potential to provide insights into the functional aspects of migratory connectivity, allowing an understanding of how connectivity affects the response of migratory populations to environmental changes and localised pressures, and how it may in turn be affected by these changes and pressures. Data integration is therefore likely to be a major tool for opening up the field of migratory connectivity in the coming years.
Availability of data and materials
Certain data included herein are derived from Clarivate Web of Science. © Copyright Clarivate 2023. All rights reserved. https://www.webofscience.com. Web of Science and Clarivate are trademarks of their respective owners and used herein with permission.
References
Åkesson S, Helm B. Endogenous programs and flexibility in bird migration. Front Ecol Evol. 2020;8:78.
Vickery JA, Ewing SR, Smith KW, Pain DJ, Bairlein F, Škorpilová J, et al. The decline of Afro-Palaearctic migrants and an assessment of potential causes. Ibis. 2014;156(1):1–22.
Sanderson FJ, Donald PF, Pain DJ, Burfield IJ, van Bommel FPJ. Long-term population declines in Afro-Palearctic migrant birds. Biol Conserv. 2006;131(1):93–105.
Rockwell SM, Wunderle JM, Sillett TS, Bocetti CI, Ewert DN, Currie D, et al. Seasonal survival estimation for a long-distance migratory bird and the influence of winter precipitation. Oecologia. 2017;183(3):715–26.
Sillett TS, Holmes RT. Variation in survivorship of a migratory songbird throughout its annual cycle. J Anim Ecol. 2002;71(2):296–308.
Senner NR, Verhoeven MA, Abad-Gómez JM, Alves JA, Hooijmeijer JCEW, Howison RA, et al. High migratory survival and highly variable migratory behavior in black-tailed godwits. Front Ecol Evol. 2019;7:96.
Norris DR, Marra PP. Seasonal interactions, habitat quality, and population dynamics in migratory birds. Condor. 2007;109(3):535–47.
Webster MS, Marra PP, Haig SM, Bensch S, Holmes RT. Links between worlds: unraveling migratory connectivity. Trends Ecol Evol. 2002;17(2):76–83.
Convention on the Conservation of Migratory Species of Wild Animals. Bonn, Germany; 1979. https://www.cms.int/.
Hostetler JA, Sillett TS, Marra PP. Full-annual-cycle population models for migratory birds. Auk. 2015;132(2):433–49.
Bauer S, Lisovski S, Hahn S. Timing is crucial for consequences of migratory connectivity. Oikos. 2016;125(5):605–12.
Rushing CS, Van Tatenhove AM, Sharp A, Ruiz-Gutierrez V, Freeman MC, Sykes PW, et al. Integrating tracking and resight data enables unbiased inferences about migratory connectivity and winter range survival from archival tags. Ornithol Appl. 2021;123(2):duab010.
Stanley CQ, McKinnon EA, Fraser KC, Macpherson MP, Casbourn G, Friesen L, et al. Connectivity of wood thrush breeding, wintering, and migration sites based on range-wide tracking. Conserv Biol. 2015;29(1):164–74.
Haig SM, Murphy SP, Matthews JH, Arismendi I, Safeeq M. Climate-Altered wetlands challenge waterbird use and migratory connectivity in arid landscapes. Sci Rep. 2019;9:4666.
Taylor CM, Stutchbury BJM. Effects of breeding versus winter habitat loss and fragmentation on the population dynamics of a migratory songbird. Ecol Appl. 2016;26(2):424–37.
Finch T, Butler SJ, Franco AMA, Cresswell W. Low migratory connectivity is common in long-distance migrant birds. J Anim Ecol. 2017;86(3):662–73.
Martin TG, Chadès I, Arcese P, Marra PP, Possingham HP, Norris DR. Optimal conservation of migratory species. PLoS ONE. 2007;2(8): e751.
Runge CA, Martin TG, Possingham HP, Willis SG, Fuller RA. Conserving mobile species. Front Ecol Environ. 2014;12(7):395–402.
Rushing CS, Rubenstein M, Lyons JE, Runge MC. Using value of information to prioritize research needs for migratory bird management under climate change: a case study using federal land acquisition in the United States. Biol Rev. 2020;95(4):1109–30.
Wood HB. The history of bird banding. Auk. 1945;62(2):256–65.
Chamberlain CP, Blum JD, Holmes RT, Feng X, Sherry TW, Graves GR. The use of isotope tracers for identifying populations of migratory birds. Oecologia. 1997;109(1):132–41.
Haig SM, Gratto-Trevor CL, Mullins TD, Colwell MA. Population identification of western hemisphere shorebirds throughout the annual cycle. Mol Ecol. 1997;6(5):413–27.
Hobson KA, Wassenaar LI. Linking breeding and wintering grounds of neotropical migrant songbirds using stable hydrogen isotopic analysis of feathers. Oecologia. 1996;109(1):142–8.
Taylor P, Crewe T, Mackenzie S, Lepage D, Aubry Y, Crysler Z, et al. The Motus Wildlife Tracking System: a collaborative research network to enhance the understanding of wildlife movement. Avian Conserv Ecol. 2017;12(1):8.
Wikelski M, Kays RW, Kasdin NJ, Thorup K, Smith JA, Swenson GW. Going wild: what a global small-animal tracking system could do for experimental biologists. J Exp Biol. 2007;210(2):181–6.
Fuller MR, Seegar WS, Schueck LS. Routes and travel rates of migrating Peregrine Falcons Falco peregrinus and Swainson’s Hawks Buteo swainsoni in the Western Hemisphere. J Avian Biol. 1998;29(4):433–40.
Stutchbury BJM, Tarof SA, Done T, Gow E, Kramer PM, Tautin J, et al. Tracking long-distance songbird migration by using geolocators. Science. 2009;323(5916):896–896.
Vansteelant WMG, Kekkonen J, Byholm P. Wind conditions and geography shape the first outbound migration of juvenile honey buzzards and their distribution across sub-Saharan Africa. Proc R Soc B. 1855;2017(284):20170387.
Jiguet F, Robert A, Lorrillière R, Hobson KA, Kardynal KJ, Arlettaz R, et al. Unravelling migration connectivity reveals unsustainable hunting of the declining ortolan bunting. Sci Adv. 2019;5(5):eaau2642.
Müller MF, Banks SC, Crewe TL, Campbell HA. The rise of animal biotelemetry and genetics research data integration. Ecol Evol. 2023;13(3): e9885.
Norevik G, Åkesson S, Artois T, Beenaerts N, Conway G, Cresswell B, et al. Wind-associated detours promote seasonal migratory connectivity in a flapping flying long-distance avian migrant. J Anim Ecol. 2020;89(2):635–46.
Thorup K, Conn PB. Estimating the seasonal distribution of migrant bird species: can standard ringing data be used? In: Thomson DL, Cooch EG, Conroy MJ, editors. Modeling demographic processes in marked populations. Boston: Springer; 2009. p. 1107–17.
Cohen EB, Hostetler JA, Royle JA, Marra PP. Estimating migratory connectivity of birds when re-encounter probabilities are heterogeneous. Ecol Evol. 2014;4(9):1659–70.
Korner-Nievergelt F, Liechti F, Thorup K. A bird distribution model for ring recovery data: where do the European robins go? Ecol Evol. 2014;4(6):720–31.
Korner-Nievergelt F, Liechti F, Hahn S. Migratory connectivity derived from sparse ring reencounter data with unknown numbers of ringed birds. J Ornithol. 2012;153(3):771–82.
Korner-Nievergelt F, Schaub M, Thorup K, Vock M, Kania W. Estimation of bird distribution based on ring re-encounters: precision and bias of the division coefficient and its relation to multi-state models. Bird Study. 2010;57(1):56–68.
Korner-Nievergelt F, Sauter A, Atkinson PW, Guélat J, Kania W, Kéry M, et al. Improving the analysis of movement data from marked individuals through explicit estimation of observer heterogeneity. J Avian Biol. 2010;41(1):8–17.
Thorup K, Korner-Nievergelt F, Cohen EB, Baillie SR. Large-scale spatial analysis of ringing and re-encounter data to infer movement patterns: a review including methodological perspectives. Methods Ecol Evol. 2014;5(12):1337–50.
Procházka P, Hahn S, Rolland S, van der Jeugd H, Csörgő T, Jiguet F, et al. Delineating large-scale migratory connectivity of reed warblers using integrated multistate models. Divers Distrib. 2017;23(1):27–40.
Von Rönn JAC, Grüebler MU, Fransson T, Köppen U, Korner-Nievergelt F. Integrating stable isotopes, parasite, and ring-reencounter data to quantify migratory connectivity—a case study with Barn Swallows breeding in Switzerland, Germany, Sweden, and Finland. Ecol Evol. 2020;10(4):2225–37.
Blackburn E, Burgess M, Freeman B, Risely A, Izang A, Ivande S, et al. Light stalks increase the precision and accuracy of non-breeding locations calculated from geolocator tags: a field test from a long-distance migrant. Bird Study. 2019;66(3):353–65.
Hallworth MT, Marra PP. Miniaturized GPS tags identify non-breeding territories of a small breeding migratory songbird. Sci Rep. 2015;5:11069.
Lisovski S, Bauer S, Briedis M, Davidson SC, Dhanjal-Adams KL, Hallworth MT, et al. Light-level geolocator analyses: a user’s guide. J Anim Ecol. 2020;89(1):221–36.
Scarpignato AL, Harrison AL, Newstead DJ, Niles LJ, Porter RR, van den Tillaart M, et al. Field-testing a new miniaturized GPS-Argos satellite transmitter (3.5 g) on migratory shorebirds. Wader Study. 2016;123(3):240–6.
DeSaix MG, Bulluck LP, Eckert AJ, Viverette CB, Boves TJ, Reese JA, et al. Population assignment reveals low migratory connectivity in a weakly structured songbird. Mol Ecol. 2019;28(9):2122–35.
Ruegg KC, Anderson EC, Paxton KL, Apkenas V, Lao S, Siegel RB, et al. Mapping migration in a songbird using high-resolution genetic markers. Mol Ecol. 2014;23(23):5726–39.
Hallworth MT, Studds CE, Scott Sillett T, Marra PP. Do archival light-level geolocators and stable hydrogen isotopes provide comparable estimates of breeding-ground origin? Auk. 2013;130(2):273–82.
Hobson KA, Wunder MB, Van Wilgenburg SL, Clark RG, Wassenaar LI. A method for investigating population declines of migratory birds using stable isotopes: origins of harvested lesser scaup in North America. PLoS ONE. 2009;4(11): e7915.
Royle JA, Rubenstein DR. The role of species abundance in determining breeding origins of migratory birds with stable isotopes. Ecol Appl. 2004;14(6):1780–8.
Hobson KA, Norris DR, Kardynal KJ, Yohannes E. Animal migration. In: Tracking animal migration with stable isotopes. Elsevier; 2019. p. 1–23.
Giunchi D, Caccamo C, Mori A, Fox JW, Rodríguez-Godoy F, Baldaccini NE, et al. Pattern of non-breeding movements by Stone-curlews Burhinus oedicnemus breeding in Northern Italy. J Ornithol. 2015;156(4):991–8.
Lisovski S, Németh Z, Wingfield JC, Krause JS, Hobson KA, Seavy NE, et al. Migration pattern of Gambel’s white-crowned sparrow along the Pacific Flyway. J Ornithol. 2019;160(4):1097–107.
Finch T, Dunning J, Kiss O, Račinskis E, Schwartz T, Sniauksta L, et al. Insights into the migration of the European Roller from ring recoveries. J Ornithol. 2017;158(1):83–90.
Bégin-Marchand C, Desrochers A, Taylor PD, Tremblay JA, Berrigan L, Frei B, et al. Spatial structure in migration routes maintained despite regional convergence among eastern populations of Swainson’s Thrushes. Mov Ecol. 2021;9(1):23.
Boulet M, Norris DR. Introduction: the past and present of migratory connectivity. Ornithol Monogr. 2006;61:1–13.
El-Arabany N, Sorensen M, Hansson B. Inferring the links between breeding and wintering grounds in a Palearctic–African migratory bird, the Great Reed Warbler, using mitochondrial DNA data. Afr Zool. 2015;50(3):241–8.
Korner-Nievergelt F, Prévot C, Hahn S, Jenni L, Liechti F. The integration of mark reencounter and tracking data to quantify migratory connectivity. Ecol Model. 2017;344:87–94.
Ambrosini R, Møller AP, Saino N. A quantitative measure of migratory connectivity. J Theor Biol. 2009;257(2):203–11.
Cohen EB, Hostetler JA, Hallworth MT, Rushing CS, Sillett TS, Marra PP. Quantifying the strength of migratory connectivity. Methods Ecol Evol. 2018;9(3):513–24.
Cohen EB, Rushing CR, Moore FR, Hallworth MT, Hostetler JA, Ramirez MG, et al. The strength of migratory connectivity for birds en route to breeding through the Gulf of Mexico. Ecography. 2019;42(4):658–69.
Smith EL, Reudink MW, Marra PP, Mckellar AE, Van Wilgenburg SL. Breeding origins and migratory connectivity at a northern roost of Vaux’s Swift, a declining aerial insectivore. Condor. 2019;121(3):duz034.
Lisovski S, Hahn S. GeoLight—processing and analysing light-based geolocator data in R. Methods Ecol Evol. 2012;3(6):1055–9.
Bennett RE, Swarthout SB, Bolsinger JS, Rodewald AD, Rosenberg KV, Rohrbaugh RW. Extreme genetic similarity does not predict non-breeding distribution of two closely related warblers. J Field Ornithol. 2017;88(2):156–68.
Hobson KA, Kardynal KJ. Western Veeries use an eastern shortest-distance pathway: new insights to migration routes and phenology using light-level geolocators. Auk. 2015;132(3):540–50.
Knight EC, Harrison AL, Scarpignato AL, Van Wilgenburg SL, Bayne EM, Ng JW, et al. Comprehensive estimation of spatial and temporal migratory connectivity across the annual cycle to direct conservation efforts. Ecography. 2021;44(5):665–79.
Lippert F, Kranstauber B, Forré PD, van Loon EE. Learning to predict spatiotemporal movement dynamics from weather radar networks. Methods Ecol Evol. 2022;13(12):2811–26.
Hallworth MT, Sillett TS, Van Wilgenburg SL, Hobson KA, Marra PP. Migratory connectivity of a neotropical migratory songbird revealed by archival light-level geolocators. Ecol Appl. 2015;25(2):336–47.
Smith ANH, Anderson MJ, Pawley MDM. Could ecologists be more random? Straightforward alternatives to haphazard spatial sampling. Ecography. 2017;40(11):1251–5.
Vickers SH, Franco AMA, Gilroy JJ. Sensitivity of migratory connectivity estimates to spatial sampling design. Mov Ecol. 2021;9(1):16.
Brlík V, Koleček J, Burgess M, Hahn S, Humple D, Krist M, et al. Weak effects of geolocators on small birds: a meta-analysis controlled for phylogeny and publication bias. J Anim Ecol. 2020;89(1):207–20.
Clegg SM, Kelly JF, Kimura M, Smith TB. Combining genetic markers and stable isotopes to reveal population connectivity and migration patterns in a Neotropical migrant, Wilson’s warbler (Wilsonia pusilla). Mol Ecol. 2003;12(4):819–30.
Contina A, Alcantara JL, Bridge ES, Ross JD, Oakley WF, Kelly JF, et al. Genetic structure of the Painted Bunting and its implications for conservation of migratory populations. Ibis. 2019;161(2):372–86.
Bay RA, Karp DS, Saracco JF, Anderegg WRL, Frishkoff LO, Wiedenfeld D, et al. Genetic variation reveals individual-level climate tracking across the annual cycle of a migratory bird. Ecol Lett. 2021;24(4):819–28.
Ruegg KC, Anderson EC, Harrigan RJ, Paxton KL, Kelly JF, Moore F, et al. Genetic assignment with isotopes and habitat suitability (gaiah), a migratory bird case study. Methods Ecol Evol. 2017;8(10):1241–52.
Bowen GJ, Wassenaar LI, Hobson KA. Global application of stable hydrogen and oxygen isotopes to wildlife forensics. Oecologia. 2005;143(3):337–48.
Graves GR, Romanek CS, Navarro AR. Stable isotope signature of philopatry and dispersal in a migratory songbird. PNAS. 2002;99(12):8096–100.
Bearhop S, Furness RW, Hilton GM, Votier SC, Waldron S. A forensic approach to understanding diet and habitat use from stable isotope analysis of (avian) claw material. Funct Ecol. 2003;17(2):270–5.
Austin RE, Wynn RB, Votier SC, Trueman C, McMinn M, Rodríguez A, et al. Patterns of at-sea behaviour at a hybrid zone between two threatened seabirds. Sci Rep. 2019;9(1):14720.
Wunder MB. Determining geographic patterns of migration and dispersal using stable isotopes in keratins. J Mammal. 2012;93(2):360–7.
Vander Zanden HB, Nelson DM, Wunder MB, Conkling TJ, Katzner T. Application of isoscapes to determine geographic origin of terrestrial wildlife for conservation and management. Biol Conserv. 2018;228:268–80.
Cardenas-Ortiz L, Bayly NJ, Kardynal KJ, Hobson KA. Defining catchment origins of a geographical bottleneck: implications of population mixing and phenological overlap for the conservation of Neotropical migratory birds. The Condor. 2020;122(2):1–13.
Hobson KA, Van Wilgenburg SL, Faaborg J, Toms JD, Rengifo C, Sosa AL, et al. Connecting breeding and wintering grounds of Neotropical migrant songbirds using stable hydrogen isotopes: a call for an isotopic atlas of migratory connectivity. J Field Ornithol. 2014;85(3):237–57.
Hobson KA, Kardynal KJ. An isotope (δ34S) filter and geolocator results constrain a dual feather isoscape (δ2H, δ13C) to identify the wintering grounds of North American Barn Swallows. Auk. 2016;133(1):86–98.
Haché S, Hobson KA, Bayne EM, Van Wilgenburg SL, Villard MA. Tracking natal dispersal in a coastal population of a migratory songbird using feather stable isotope (δ2H, δ34S) tracers. PLoS ONE. 2014;9(4): e94437.
Milano S, Frahnert S, Hallau A, Töpfer T, Woog F, Voigt CC. Isotope record tracks changes in historical wintering ranges of a passerine in sub-Saharan Africa. Glob Change Biol. 2021;00:1–9.
Conklin JR, Battley PF, Potter MA, Ruthrauff DR. Geographic variation in morphology of Alaska-breeding bar-tailed godwits (Limosa lapponica) is not maintained on their nonbreeding grounds in New Zealand. Auk. 2011;128(2):363–73.
Szép T, Hobson KA, Vallner J, Piper SE, Kovács B, Szabó DZ, et al. Comparison of trace element and stable isotope approaches to the study of migratory connectivity: an example using two hirundine species breeding in Europe and wintering in Africa. J Ornithol. 2009;150(3):621.
Somveille M, Bay RA, Smith TB, Marra PP, Ruegg KC. A general theory of avian migratory connectivity. Ecol Lett. 2021;24(9):1848–58.
Vincent JG, Schuster R, Wilson S, Fink D, Bennett JR. Clustering community science data to infer songbird migratory connectivity in the Western Hemisphere. Ecosphere. 2022;13(4): e4011.
Fuentes M, Van Doren BM, Fink D, Sheldon D. BirdFlow: learning seasonal bird movements from eBird data. Methods Ecol Evol. 2023;14(3):923–38.
Smith RB, Meehan TD, Wolf BO. Assessing migration patterns of sharp-shinned hawks Accipiter striatus using stable-isotope and band encounter analysis. J Avian Biol. 2003;34(4):387–92.
Macdonald CA, Fraser KC, Gilchrist HG, Kyser TK, Fox JW, Love OP. Strong migratory connectivity in a declining arctic passerine. Anim Migr. 2012;1(1):23–30.
Boulet M, Gibbs HL, Hobson KA. Integrated analysis of genetic, stable isotope, and banding data reveal migratory connectivity and flyways in the Northern Yellow Warbler (Dendroica petechia; Aestiva group). Ornithol Monogr. 2006;61:29–78.
Van Wilgenburg SL, Hobson KA. Combining stable-isotope (δD) and band recovery data to improve probabilistic assignment of migratory birds to origin. Ecol Appl. 2011;21(4):1340–51.
Chabot AA, Hobson KA, Van Wilgenburg SL, McQuat GJ, Lougheed SC. Advances in linking wintering migrant birds to their breeding-ground origins using combined analyses of genetic and stable isotope markers. PLoS ONE. 2012;7(8): e43627.
Pekarsky S, Angert A, Haese B, Werner M, Hobson KA, Nathan R. Enriching the isotopic toolbox for migratory connectivity analysis: a new approach for migratory species breeding in remote or unexplored areas. Divers Distrib. 2015;21(4):416–27.
Rushing CS, Ryder TB, Saracco JF, Marra PP. Assessing migratory connectivity for a long-distance migratory bird using multiple intrinsic markers. Ecol Appl. 2014;24(3):445–56.
Rundel CW, Wunder MB, Alvarado AH, Ruegg KC, Harrigan R, Schuh A, et al. Novel statistical methods for integrating genetic and stable isotope data to infer individual-level migratory connectivity. Mol Ecol. 2013;22(16):4163–76.
Abadi F, Gimenez O, Ullrich B, Arlettaz R, Schaub M. Estimation of immigration rate using integrated population models. J Appl Ecol. 2010;47(2):393–400.
Rushing CS, Marra PP, Studds CE. Incorporating breeding abundance into spatial assignments on continuous surfaces. Ecol Evol. 2017;7(11):3847–55.
McKinnon EA, Stanley CQ, Fraser KC, MacPherson MM, Casbourn G, Marra PP, et al. Estimating geolocator accuracy for a migratory songbird using live ground-truthing in tropical forest. Anim Migr. 2013;2012(2):1–8.
Contina A, Magozzi S, Vander Zanden HB, Bowen GJ, Wunder MB. Optimizing stable isotope sampling design in terrestrial movement ecology research. Methods Ecol Evol. 2022;13(6):1237–49.
González-Prieto AM, Hobson KA, Bayly NJ, Gómez C. Geographic Origins and timing of fall migration of the Veery in Northern Colombia. Condor. 2011;113(4):860–8.
Hill JM, Renfrew RB. Migratory patterns and connectivity of two North American grassland bird species. Ecol Evol. 2018;9(1):680–92.
Meehan TD, Saunders SP, DeLuca WV, Michel NL, Grand J, Deppe JL, et al. Integrating data types to estimate spatial patterns of avian migration across the Western Hemisphere. Ecol Appl. 2022;32(7): e2679.
Zipkin EF, Zylstra ER, Wright AD, Saunders SP, Finley AO, Dietze MC, et al. Addressing data integration challenges to link ecological processes across scales. Front Ecol Environ. 2021;19(1):30–8.
Reese J, Viverette C, Tonra CM, Bayly NJ, Boves TJ, Johnson E, et al. Using stable isotopes to estimate migratory connectivity for a patchily distributed, wetland-associated Neotropical migrant. Condor. 2019;121(4):1–15.
Pacifici K, Reich BJ, Miller DAW, Gardner B, Stauffer G, Singh S, et al. Integrating multiple data sources in species distribution modeling: a framework for data fusion. Ecology. 2017;98(3):840–50.
Marra PP, Hobson KA, Holmes RT. Linking winter and summer events in a migratory bird by using stable-carbon isotopes. Science. 1998;282(5395):1884–6.
Culp LA, Cohen EB, Scarpignato AL, Thogmartin WE, Marra PP. Full annual cycle climate change vulnerability assessment for migratory birds. Ecosphere. 2017;8(3):e01565.
Knight SM, Bradley DW, Clark RG, Gow EA, Bélisle M, Berzins LL, et al. Constructing and evaluating a continent-wide migratory songbird network across the annual cycle. Ecol Monogr. 2018;88(3):445–60.
Hallworth MT, Bayne E, McKinnon E, Love O, Tremblay JA, Drolet B, et al. Habitat loss on the breeding grounds is a major contributor to population declines in a long-distance migratory songbird. Proc R Soc B. 2021;288:20203164.
Taylor CM, Norris DR. Population dynamics in migratory networks. Theor Ecol. 2010;3(2):65–73.
Marra PP, Cohen EB, Loss SR, Rutter JE, Tonra CM. A call for full annual cycle research in animal ecology. Biol Lett. 2015;11(8):20150552.
Acknowledgements
We are particularly grateful to Paul Clémençon and Marc Kéry whose comments on the early stages of the manuscript were invaluable, Jan Perret who provided methods and advice for the analysis of the Web of Science data, and two anonymous reviewers whose valuable comments helped to improve this manuscript.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Author information
Authors and Affiliations
Contributions
KAG reviewed the existing literature and wrote the manuscript; AB, CF and FJ contributed substantially to revisions and gave their final approval for publication. All authors read and approved the final manuscript.
Corresponding author
Ethics declarations
Ethics approval and consent to participate
Not applicable.
Consent for publication
Not applicable.
Competing interests
The authors declare that they have no competing interests.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Additional file 1
. Filtering of publications from the Web of Science: methodology for searching and analysing the Web of Science database for what concerns migratory connectivity studies.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
About this article
Cite this article
Gregory, K.A., Francesiaz, C., Jiguet, F. et al. A synthesis of recent tools and perspectives in migratory connectivity studies. Mov Ecol 11, 69 (2023). https://doi.org/10.1186/s40462-023-00388-z
Received:
Accepted:
Published:
DOI: https://doi.org/10.1186/s40462-023-00388-z