Box 1 What description of migratory connectivity?

Studies tackling migratory connectivity issues revolve around at least one out of three features that describe the movements of individuals. Concepts are described below for spring migration (movements from breeding to non-breeding/wintering sites), but they are equally applicable to autumn migration (movements from non-breeding/wintering to breeding sites).

(i) Connectivity patterns:

• What?

Qualitative description of links between populations or sites at different stages of the migratory cycle. Patterns can be further described based on how the links are arranged in relation to each other (“longitudinal”/ “parallel”/ “leap-frog”/ “cross-wise” migration; see [55] for a review).

• How?

Generally based solely on the spatial assignment of samples, individuals or groups of individuals to their location in the previous season.

• What data?

All approaches used to track individuals through space (banding, tracking technologies, genetic and isotopic assignments, e.g. [52, 56]).

(ii) Transition probabilities

• What?

Probability that an individual of population A moves to population B. Transition probabilities can be thought of as a quantitative description of links between populations, thus giving a direct estimation of migratory connectivity if estimated between breeding and wintering populations [57]. This concept ties the population-level metrics of migratory connectivity with individual-level behaviour.

• How?

Relies on models that are able to account for biases due to uneven sampling that obscure the true transition probabilities [12, 32, 57].

• What data?

Such models have mostly been developed for banding data [32, 34], and more recently tracking [12, 57] and isotope data [40].

(iii) Connectivity strength

• What?

A metrics describing how much the structuring of populations in one season is conserved in the next season [8]. There is a continuum between a complete mixing of previously isolated populations (“weak” or “diffuse” connectivity) and the conservation of the same structure between both seasons (“strong” connectivity), which depends on both population spread and inter-population mixing [16]. This property of the migratory system is likely to have an influence on the susceptibility of species to habitat loss, climate change, and population declines [16].

• How?

Two metrics have been adapted or developed to estimate connectivity strength:

   - Mantel test quantifies the correlation between matrices of inter-individual distances (usually one for the breeding and one for the non-breeding season; [58]).

   - MC completes the Mantel test by including transition probabilities and taking biases in data collection into account [59].

• What data?

Usually performed on banding or tracking data that locates individuals with good spatial resolution, but recently adapted to include isotopic data [60, 61]. MC is notably under constant improvement to integrate more diverse data.