Post provided by Emily Cohen
The seasonal long-distance migration of all kinds of animals – from whales to dragonflies to amphibians to birds – is as astonishing a feat as it is mysterious and this is an especially exciting time to study migratory animals. In the past 20 years, rapidly advancing technologies – from tracking devices, to stable isotopes in tissues, to genomics and analytical techniques for the analysis of ring re-encounter databases – mean that it’s now possible to follow many animals throughout the year and solve many of the mysteries of migration.
What is Migratory Connectivity?
One of the many important things we’re now able to measure is migratory connectivity, the connections of migratory individuals and populations between seasons. There are really two components of migratory connectivity:
- Linking the geography of where individuals and populations occur between seasons.
- The extent, or strength, of co-occurrence of individuals and populations between seasons.
Our article ‘Quantifying the strength of migratory connectivity’ is about the strength of migratory connectivity, the distributions of populations between seasons. So, for example, strong (or high or tight) connectivity occurs when populations remain together between seasons. Weak (or low or diffuse) migratory connectivity occurs when populations don’t remain together between seasons. And, of course, in the real world we see all kinds of variation between those two extremes.
Consistent Terminology and Comparable Methods
Researchers have often reported migratory connectivity in qualitative terms, such as ‘high,’ ‘weak,’ and ‘moderate,’ but we’re not necessarily talking about the same thing with those words. This qualitative terminology, as well as the use of differing methods, has made it really difficult to advance the field and have a more cohesive and comprehensive conversation. But the events and conditions that populations are exposed to throughout the year have consequences for many biological processes, including migration schedules, reproductive success, survival and abundance, range limits, and natal dispersal. So, quantitative measurement of migratory connectivity strength is key to understanding the ecology and evolution of migratory species and developing effective conservation strategies for them.
The goal of this paper and the R package MigConnectivity is to facilitate quantitative exploration of the strength of migratory connectivity with methodology that is comparable across species, studies, and data types. Comparable measures of the strength of migratory connectivity need to be independent of species, range sizes, and data type; account for relative abundance among regions; and incorporate sampling error.
For example, most studies have not sampled migratory animals across the full extent of a species’ range, nor have they sampled in proportion to regional abundance. We describe approaches to explicitly address uneven sampling and incorporate uncertainty in estimates of migratory connectivity strength from multiple sources of sampling and process error.
We also define terminology, review the methods currently in use and build on these to increase applicability. The most commonly used metric of migratory connectivity strength is the Mantel correlation. However, this method doesn’t really work when you have transition probabilities between populations instead of locations of individuals. An extension of this method, MC, incorporates population-specific transition probabilities that can be derived from many data types and accounts for uneven sampling. We included a decision support tree in our paper for use of methods to quantify the strength of migratory connectivity (given available data) as well.
What’s Next for Us with this Work?
The R package was built with the objective of increasing collaboration between scientists studying migratory animals around the world. Right now at the Smithsonian Migratory Bird Center, we’re working on incorporating these methods into many ongoing projects including range-restricted species such as the endangered Kirtland’s Warbler and species with large geographic distributions such as the Prothonotary Warbler and Gray Catbird, among others. We’re also working on an analysis of how migratory connectivity strength influences population viability and an Atlas of Migratory Connectivity for North America’s breeding birds.
To find out more, read our Methods in Ecology and Evolution article ‘Quantifying the strength of migratory connectivity’ and download the R package
Watch for package updates and contact Jeff Hostetler or Michael Hallworth with questions and potential package additions.
Methods Blog 
One thought on “Sticking Together or Drifting Apart? Quantifying the Strength of Migratory Connectivity”