Gauge-and-compass migration: inherited magnetic headings and signposts can adapt to changing geomagnetic landscapes

Abstract

Background: For many migratory species, inexperienced (naïve) individuals reach remote non-breeding areas independently using one or more inherited compass headings and, potentially, magnetic signposts to gauge where to switch between compass headings. Inherited magnetic-based migration has not yet been assessed as a population-level process, particularly across strong geomagnetic gradients or where long-term geomagnetic shifts (hereafter, secular variation) could create mismatches with magnetic headings. Therefore, it remains unclear whether inherited magnetic headings and signposts could potentially adapt to secular variation under natural selection.

Methods: To address these unknowns, we modelled migratory orientation programs using an evolutionary algorithm incorporating global geomagnetic data (1900-2023). Modelled population mixing incorporated both natal dispersal and trans-generational inheritance of magnetic headings and signposts, including intrinsic (stochastic) variability in inheritance. Using the model, we assessed robustness of trans-hemispheric migration of a migratory songbird whose Nearctic breeding grounds have undergone rapid secular variation (mean 34° clockwise drift in declination, 1900-2023), and which travels across strong geomagnetic gradients via Europe to Africa.

Results: Model-evolved magnetic-signposted migration was overall successful throughout the 124-year period, with 60-90% mean successful arrival across a broad range in plausible precision in compass headings and gauging signposts. Signposted migration reduced trans-Atlantic flight distances and was up to twice as successful compared with non-signposted migration. Magnetic headings shifted plastically in response to the secular variation (mean 16°-17° among orientation programs), whereas signpost latitudes were more constrained (3°-5° mean shifts). This plasticity required intrinsic variability in inheritance (model-evolved σ ≈ 2.6° standard error), preventing clockwise secular drift from causing unsustainable open-ocean flights.

Conclusions: Our study supports the potential long-term viability of inherited magnetic migratory headings and signposts, and illustrates more generally how inherited migratory orientation programs can both mediate and constrain evolution of routes, in response to global environmental change.

Keywords: Animal migration; Bet-hedging; Evolutionary strategy algorithm; Geomagnetic core field; Magnetic compass; Migratory orientation program; Natal dispersal; Northern wheatear; Secular variation; Zugknick.

Fig. 1

Extreme geomagnetic gradients and temporal shifts, illustrated for bird migration across the North Atlantic. a Contours of mean-field geomagnetic declination in October 2010 (degrees clockwise from true to magnetic North; colour scale on right), with modelled natal range (in brown) and wintering grounds (goal area; green) of northern wheatears (Oenanthe oenanthe leucorhoa). The brown dashed line and arrow depicts the approximate actual route from Iqaluit, Baffin Island taken by a juvenile (leucorhoa) northern wheatear tagged in 2010 using light-level geolocation [34]. Distance-minimizing great circle routes are straight lines in the stereographic projection [35]. b, c Contours of temporal change (secular variation) in mean-field declination from b 1900–1960 and c 1960–2023, with colour scale on right. Magnetic data are from a global IGRF modelled data of the Earth’s core-field [36, 37]

Fig. 2

Modellled parent selection and trait inheritance (mixing) among successful migrants. a For each breeding location (orange asterisk) in each modelled year, two among all successful migrants were selected based on a sampled distance, $d^{\ast }$, between their natal and the focal breeding location relative to the population mean dispersal, D_N (illustrated as orange circle, with trajectories and sampled dispersal distances of two candidate “parents” depicted in blue and green). Selection probability followed a half-normal distribution (lower left graph, equation, top right). D_N was “evolved” as an individual trait during the initial model “spin-up” (see Additional file 2). b Next-generation (offspring) inheritance of migratory headings, ${\theta }_1^{\ast }$, ${\theta }_2^{\ast }$ and signposts, $s^{\ast }$, were sampled from circular von Mises distributions (CN, for headings and inclination signposts) and normal distributions (N, for intensity signposts), centred around between-parental means, together with intrinsic variability in inheritance for each trait (the latter also “evolved” during model spin-up; see Additional file 2). For non-signposted migration, only the first heading, ${\theta }_1^{\ast }$, is inherited

Fig. 3

Model-evolved migratory trajectories and arrival success of inexperienced (naïve) leucorhoa wheatears (see Fig. 1). Randomly-sampled predicted trajectories from 2023, colour-coded to flight direction (degrees clockwise from geographic N) based on a non-signposted migration, following a constant inherited magnetic heading, b a magnetic signpost based on inclination and c a signpost based on geomagnetic intensity. The results in a–c are with default model parameters (see “Methods” section and Table 1). For b, c, encountering a signpost (inherited threshold geomagnetic value) triggers a shift to a second model-evolved inherited heading. Successful arrival in Africa is indicated by white circles, and pink tracks represent unsuccessful individuals. Straight lines represent great circle routes in the stereographic azimuthal projection; d arrival success (percentage of population) for non-signposted (solid orange line), inclination-signposted (solid blue line) and intensity-signposted migration (solid green line). Dashed lines depict success when the model is parameterised by geomagnetic data in reverse chronological order (2023–1900)

Fig. 4

Sensitivity of modelled magnetic-based migration to precision among flight-steps and in gauging signposts. All panels depict long-term mean arrival success (%) and standard deviation among years for non-signposted migration (orange dashed line), and signposted migration based on inclination (dot-dashed blue line) and intensity (solid green line). Arrival success is plotted as a function of a precision among flight-steps (degrees), b precision in gauging signposts based on inclination (degrees, dot-dashed blue line), and based on intensity (percent intensity, solid green line), and c mean natal dispersal (i.e., distance between breeding and natal grounds, km). Circle symbols depict default parameters of a 15° flight-flight-step precision and b 5° precision in detection of signpost inclination and 2% of intensity signpost. The star symbols in c depict (default) model-evolved mean natal dispersal (as in Fig. 3)

Fig. 5

Evolution of modelled intensity-signposted migration of leucoroha wheatears to geomagnetic secular variation. Coloured symbols of 5000 randomly-selected successful modelled individuals illustrate a, b inherited magnetic headings (clockwise degrees from magnetic N) and c, d Zugknick latitudes (degrees), from 1900 (a, c) and 2023 (b, d). Yearly changes relative to 1900 regarding e initial inherited headings (° clockwise) and f intensity signposts (nT) among randomly-selected individuals (natal locations), as a function of change in declination at the natal site (° clockwise) since 1900, colour-coded per year (scale on top). Orange dashed lines represent mean changes in e headings and f intensity signposts (nT), sorted in 5° bins of declination change. a–d Stereographic azimuthal projection

Fig. 6

Intrinsic variability in inheritance facilitates robustness of magnetic headings and signposts to geomagnetic secular variation. Circles represent mean successful arrival (left axes) and triangles mortality over water (right axes) among intensity-signposted migrants (Fig. 3c), with symbol colours depicting population-mean changes in inherited headings since 1900 (circles, clockwise degrees from magnetic N) and in signpost magnitude (triangles, nT). a With model-evolved standard deviation in inherited headings (2.6°) and intensity signposts (0.53%); b as a but with an inherited geographic (star) primary and in-flight compass; c is as a and d as in b, but without intrinsic standard deviations in inheritance of headings and signposts