Module 8: Climate & Migration Shifts

Marcel Visser and Christiaan Both (Proc. R. Soc. B, 2005) formalised the trophic-mismatch hypothesis: long-distance migrants whose breeding habitat warms rapidly must advance their arrival date to remain phenologically synchronised with their prey. The challenge is that the cues initiating migration—day length at the non-breeding ground, endogenous clock—do not respond to temperature changes at the breeding ground. Caterpillar emergence, bud-burst, and peak insect abundance, by contrast, respond directly to spring temperature. The result: resources advance faster than migrant arrival, producing a growing phenological gap.

Gordo (2007, Climate Research) meta-analysed 296 time series of European migrant bird arrival dates and documented arrival-date advances of 2–4 days per decade. Meanwhile, oak-budburst and winter-moth-caterpillar peak dates in the same region advance at 4–7 days per decade, driven more strongly by warming. The decadal differential of ~3 days per decade accumulates to a 20 +-day mismatch over a half-century—enough to push chick-feeding demand off the resource peak entirely in species with sharp prey phenology.

Christiaan Both et al. (Nature, 2006) exploited a natural experiment across 9 pied flycatcher (Ficedula hypoleuca) populations in the Netherlands that differ in the severity of seasonal warming. Populations breeding in habitats where peak caterpillar biomass had advanced most strongly had declined by ~90% over two decades, whereas populations where caterpillar timing remained relatively stable declined by less than 10%. The per-population decline was linearly correlated with the degree of phenological mismatch, providing one of the strongest causal tests of the mismatch hypothesis to date.

The underlying mechanism is chick starvation: when the peak caterpillar biomass falls before nestlings are old enough to benefit, parents return to the nest with smaller and fewer prey items, per-nest fledgling survival drops, and annual recruitment falls below the replacement level. Pied flycatchers are Afrotropical migrants and their arrival dates are cued by a photoperiodic signal in the African non-breeding ground, which is not responsive to European warming—they are locked into a timing window that was adaptive a century ago but is now misaligned.

The East Asian–Australasian Flyway carries 50 million shorebirds annually from Siberia and Alaska to Australia and New Zealand via the Yellow Sea. For species like the bar-tailed godwit and red knot the Yellow Sea intertidal mudflats are the sole refuelling stopover between the breeding ground and the non-breeding ground. The mudflats have been under intense reclamation pressure since the 1980s: land-reclamation projects in China and South Korea have converted roughly 65% of Yellow Sea intertidal habitat to industrial, agricultural, and urban land over 30 years (Studds et al., Nature Communications, 2017).

Studds et al. combined bird-tracking data, site-specific population counts, and satellite-derived habitat-loss maps for 16 shorebird species and showed that each species’ decline rate was linearly predicted by its reliance on the lost sites. Red knot (Calidris canutus rogersi) has declined by ~58%, bar-tailed godwit (Limosa lapponica baueri) by ~45%, great knot by ~56%—in each case with a slope consistent with the amount of Yellow Sea habitat used. The Yellow Sea’s 65% site-area loss is thus the dominant driver of the flyway-wide shorebird decline, as predicted by a simple linear relationship between stopover-area availability and migrant survival.

Global warming is reshaping the large-scale circulation of the atmosphere. Loonin et al. (2023, Nature Climate Change) used 60 years of satellite and reanalysis data combined with IPCC CMIP6 ensembles to project that jet-stream meandering (the north-south excursions of the polar jet) will increase by 2060 under SSP2-4.5, raising the average tailwind variability experienced by long-distance migrants. The authors estimated the consequent increase in per-migration energy cost at ~9% for transatlantic passerines and ~12% for trans-Saharan species by 2060, with larger uncertainties thereafter.

These headwinds combine with the mismatch effects of sections 1–2 to threaten species already near the upper limit of their fat-carrying capacity (Pennycuick model, Module 7). A 9% cost increase is equivalent to requiring ~8 % more fat deposition at the same departure timing—unreachable for species already at hyperphagic maximum. The consequence is shorter flights, more stopovers, and greater reliance on intermediate staging habitat—precisely the habitat that is itself being lost.

The eastern monarch butterfly population has collapsed by ~84% from 1996–2014 and remains at ~20–30% of its historical carrying capacity. The drivers are two-fold: breeding-ground milkweed loss, and overwintering-habitat vulnerability. Pleasants & Oberhauser (2013; see Module 5) documented a 58% decline in midwestern milkweed from 1999–2012, tightly correlated with the introduction of glyphosate-tolerant crops. Inamine et al. (Oikos, 2016) meta-analysed 20 years of citizen-science data and confirmed that breeding-ground factors, not overwintering mortality, dominate the population trend. In December 2022 the IUCN Red List formally classified the migratory eastern monarch as Endangered.

Oyamel fir climate vulnerability

Sáenz-Romero et al. (2012, Forest Ecology and Management) projected that by 2090 the climate envelope of Abies religiosa will retreat upslope by 600–900 m and essentially disappear from its current range—a projection updated in 2020 under SSP5-8.5 to an even more dramatic loss of habitable area. Even under optimistic SSP1-2.6 mitigation scenarios, only ~30% of current oyamel-suitable area remains climatically appropriate by 2100. Assisted upslope migration of seedlings is underway but mature-forest conditions require a century to develop, creating a clear timing gap between loss of lowland habitat and availability of upland replacement.

Multi-species monitoring confirms that advance trends are widespread across avian migrants. In a dataset of 10 common European passerines and shorebirds tracked over 1970–2020, arrival-date trends cluster between −1.5 and −5.0 days per decade, with a median of −3.0 d/decade. Long-distance migrants that winter south of the Sahara advance slightly less than short-distance migrants that winter within Europe—consistent with the hypothesis that photoperiodic cueing constrains the response of trans-Saharan migrants more tightly than that of short-distance migrants, which respond directly to European temperature.

Altitudinal migrants—species that move up and down mountain slopes with the seasons rather than across latitude—show analogous responses to warming. Studies in the Peruvian Andes (Freeman & Class-Freeman, 2014) and on Mount Kinabalu (Malaysia) have documented 30–60 m per decade upslope shifts in bird community composition, with lowland specialists disappearing from mid-elevation sites and highland specialists retreating to the summits. In many cases the summit species have nowhere higher to go: species like the Rufous-capped thornbill (Ecuador) now occupy only the highest summit of their former range and face local extinction as those summits warm.

The same pattern applies to North American montane species: the dark-eyed junco breeds progressively higher each decade in the Sierra Nevada; the grey-crowned rosy-finch is at risk of losing its alpine tundra breeding habitat. For long-distance migrants that rely on alpine tundra as a stopover (broad-tailed hummingbird, Lincoln’s sparrow), the shrinking alpine belt is a further pressure.

Why do migrants not simply advance their arrival dates to track resources? The answer lies in the cueing system. Photoperiod (day-length) at the non-breeding ground is the primary trigger for migratory departure in most long-distance species. Photoperiod is locked to Earth’s orbit, not to climate, and so cannot shift with warming. Short-distance migrants can partially respond by relying on temperature cues at the non-breeding ground, but trans-Saharan and trans-equatorial migrants cannot. The result is a constrained response: arrival dates advance, but slowly, and the advance is ultimately limited by the genetic architecture of the cueing system.

Microevolutionary change can, in principle, overcome the constraint. Berthold (1992) showed that blackcap (Sylvia atricapilla) winter-range direction evolved in 30 years from mostly-Iberia to largely-UK, under artificial provision of feeders in the UK. Similar rapid evolution could in principle shift arrival dates. However, for rare and declining species with small effective population sizes, the pace of genetic change is expected to be slower than the pace of climate warming, leading to persistent mismatch and population decline.

Migratory species present a unique conservation challenge: each individual depends on habitat quality and climate conditions along the full annual cycle, and loss of any link (breeding, stopover, non-breeding) can doom the whole migration. Effective conservation therefore requires flyway-scale protection—coordinated action across national boundaries for species whose range spans continents.

• East Asian-Australasian Flyway Partnership: 37 network sites protected under formal agreements across 19 countries; successful in halting some Yellow Sea reclamation projects (e.g., Saemangeum decommissioning in South Korea).
• North American Monarch Joint Venture: coordinates milkweed-restoration planting across 16 US states and Canadian provinces; established 2 million acres of pollinator habitat by 2022.
• Mexican Monarch Biosphere Reserve: 56 000 ha legally protected since 2000, UNESCO World Heritage 2008; illegal-logging enforcement remains uneven.
• Assisted migration of oyamel fir: experimental upslope transplants in progress at 300–500 m above current range edges to establish future overwintering habitat.
• Citizen-science monitoring: eBird, Journey North, Monarch Watch, BirdTrack, Swedish Bird Survey provide high-spatial-resolution data on arrival-date trends, stopover use, and population counts.

Habitat corridors and connectivity

Landscape-scale corridor planning recognises that migrants require functionally connected stopover habitat, not just spatially distributed patches. Targeted restoration of wetland complexes (prairie-pothole region of the US and Canada), intertidal mudflats (Yellow Sea, Wadden Sea), and riparian woodlots can restore flyway connectivity where it has been most eroded. Economic-valuation studies suggest that such restoration yields benefits 3–5× investment through carbon sequestration, fisheries productivity, and tourism—in addition to migratory-biodiversity benefits.

For monarchs, the analogue of bird-caterpillar mismatch is the monarch-milkweed phenology relation. As the super-generation monarch moves northward in spring, it tracks the emergence of common milkweed shoots from Texas to Ontario at approximately 200 km per generation. Warming shifts milkweed emergence earlier in the south but also lengthens the active growing season in the north. The net effect is a compressed northward corridor with potential mismatch at the trailing (southern) edge and extended breeding opportunity at the leading (northern) edge—but with uncertainty about whether the supergeneration has the plasticity to adjust its arrival timing at overwintering sites in the face of an altered Mexican winter climate.

Beyond phenological shifts, climate change is altering the amplitude and regularity of migration itself. Irruptive species such as snowy owl, redpoll, common crossbill, and pine siskin migrate southward in variable numbers each winter depending on northern food availability (lemmings, cone crops). Climate change disrupts the lemming cycle through altered snow cover, and altered conifer cone crop cycles through temperature and precipitation anomalies. The result is more frequent, less synchronised irruptions, with southward movements of birds that might historically have remained on breeding range all winter.

At the opposite end of the spectrum, some species have become shorter-distance migrants or even residents. Purple martin, American robin, and song sparrow populations in the northeast US have progressively shortened their winter range northward over 50 years. The blackcap (Sylvia atricapilla) in continental Europe has evolved a new UK-wintering migration route (section 9bb). These changes are responses to milder winters that make shorter migrations or even residency energetically viable.

High-Arctic breeding shorebirds face a distinct form of phenological mismatch. The Arctic summer is warming 2–3× faster than the global mean, with snowmelt advancing at 4–6 days per decade. Arthropod (primarily chironomid midge) emergence responds directly to snowmelt and tracks the change tightly. The shorebirds, however, whose arrival dates are set primarily by photoperiod at wintering grounds 10–15 000 km to the south, cannot keep up. For red knot, semipalmated sandpiper, dunlin, and ruddy turnstone, measurements at long-term monitoring sites (Zackenberg, Greenland; Churchill, Manitoba; Barrow, Alaska) show progressive accumulation of hatching-vs-prey-peak mismatch of 1–2 days per decade.

Perhaps most strikingly, Saino et al. (2011) and van Gils et al. (Science, 2016) reported that red knot chicks hatched during years of advanced snowmelt showed reduced body size at fledging, reduced wing length, and reduced juvenile survival the following winter. The reduced adult body size in turn affects competitive ability on tropical wintering grounds, closing a feedback loop that propagates Arctic climate change into the annual cycle thousands of kilometres away.

Climate warming also reshapes disease pressure on migrants. The West Nile virus invasion of North America since 1999 has killed millions of wild birds, and the northward range expansion of its mosquito vectors is climate-driven. Avian malaria (Plasmodium) prevalence in migratory passerines has risen by a factor of two in some European populations over three decades. Tick-borne pathogens (Borrelia, Babesia) track tick ranges, which are advancing northward at up to 50 km per year. Each new disease introduced to a naive migratory population represents an additional selection pressure that is, like trophic mismatch, operating faster than genetic adaptation can catch up.

For monarchs the analogous pressure is the OE parasite (see Module 5). In the migratory eastern population OE prevalence is held to 5–15% by the migratory cull, but in non-migratory populations on tropical milkweed OE can exceed 70%. Climate change may facilitate the spread of the non-migratory, high-parasite-load phenotype into previously-migratory ranges, undermining the migratory-escape benefit.

Against the background of phenological constraint, some migratory species are showing measurable microevolution on decadal timescales. The iconic case is the Central European blackcap (Sylvia atricapilla), which historically migrated south-west to Iberia but now increasingly migrates north-west to the UK, where garden feeders sustain winter populations. Berthold et al. (Nature, 1992) showed that the direction is heritable, and Rolshausen et al. (Curr. Biol., 2009) showed genetic divergence between UK-overwintering and Iberia-overwintering blackcaps after only 30 generations. The evolution was facilitated by strong selection differential (higher survival in UK) and by the large effective population size of the species.

Similarly, the southward-breeding range expansion of the Eurasian collared dove from Turkey to Britain between 1930 and 1970 reflects rapid microevolution of colonising behaviour, and the reduced migration distance of some North American swallow species over recent decades points to plasticity that may underpin slower microevolutionary change. The general pattern is that migratory traits can evolve rapidly when selection is strong enough, but the pace of ongoing climate change may outstrip the response for many species.

Traditional bird conservation focused heavily on breeding-ground habitat protection. Recent research emphasises that migratory species are vulnerable throughout the full annual cycle, including the wintering (non-breeding) ground, which for many species makes up 50–80% of the year. Marra et al. (Science, 1998) established the principle of migratory connectivity—that specific breeding populations tend to overwinter in specific non-breeding regions, so the fate of a breeding population depends on conditions at its particular non-breeding range.

Using stable-isotope and genetic tracking, researchers have mapped connectivity patterns for dozens of species. Some show strong connectivity (the Swainson’s thrush breeding in Alaska winters in Costa Rica; the ones breeding in Ontario winter in Ecuador), while others show weak connectivity (breeding populations mix fully on the non-breeding ground). Strong-connectivity species are most vulnerable to habitat loss on localised non-breeding grounds; weak-connectivity species are more buffered. Conservation planning must respect connectivity patterns, protecting the non-breeding range most used by the populations of greatest concern.

The bird-and-monarch migrations examined across this course have been honed by tens of thousands of generations of selection, with molecular clocks, magnetic compasses, flight muscles, fat stores, and annual cycles tuned to particular configurations of latitude, photoperiod, and resource phenology. The configurations are now changing faster than the migrations can accommodate. Trophic mismatch, stopover-site loss, altered wind regimes, and habitat-envelope shifts are not independent stressors but interlocking pressures; each reinforces the others.

The future of long-distance migration is, nonetheless, not uniformly bleak. Species with large populations and strong phenotypic plasticity will persist; citizen-science monitoring gives unprecedented visibility into ongoing changes; targeted conservation interventions have demonstrable benefits; international agreements have begun to protect critical stopover habitat. The task of the coming century is to hold the network together long enough for genetic adaptation to catch up with climate. The science of bird and monarch migration will remain central to that task—a laboratory for measuring the pace of biospheric change and the resilience of evolved adaptations in a warming world.

This course has traced bird and monarch migration from celestial navigation at the quantum level (radical-pair magnetoreception, Module 2) through multi-stage biochemistry (cryptochromes, cardenolide resistance, hyperphagia, fat oxidation) to continental-scale ecology (trophic mismatch, flyway networks, oyamel climate-zone shifts). The breadth of the field reflects its subject: migration is simultaneously a neurobiological, biochemical, physiological, ecological, and evolutionary phenomenon. No single level of analysis tells the whole story.

Looking ahead, three research frontiers stand out. (i) The molecular identity of the magnetoreceptor—still unresolved 25 years after the radical-pair hypothesis was first proposed—is under active investigation with single-molecule biophysics. (ii) The genomic basis of migratory direction and distance, now accessible through population genomics of sister populations that differ in migration strategy, may yield the genes underlying Zugunruhe heritability. (iii) Conservation genomics of declining migrants, combining isotope tracking, citizen-science monitoring, and microevolutionary modelling, may allow targeted interventions at the populations and life-history stages most at risk.