Long-distance movement dynamics shape host microbiome richness and turnover

Abstract Host-associated microbial communities are shaped by host migratory movements. These movements can have contrasting impacts on microbiota, and understanding such patterns can provide insight into the ecological processes that contribute to community diversity. Furthermore, long-distance movements to new environments are anticipated to occur with increasing frequency due to host distribution shifts resulting from climate change. Understanding how hosts transport their microbiota with them could be of importance when examining biological invasions. Although microbial community shifts are well-documented, the underlying mechanisms that lead to the restructuring of these communities remain relatively unexplored. Using literature and ecological simulations, we develop a framework to elucidate the major factors that lead to community change. We group host movements into two types—regular (repeated/cyclical migratory movements, as found in many birds and mammals) and irregular (stochastic/infrequent movements that do not occur on a cyclical basis, as found in many insects and plants). Ecological simulations and prior research suggest that movement type and frequency, alongside environmental exposure (e.g. internal/external microbiota) are key considerations for understanding movement-associated community changes. From our framework, we derive a series of testable hypotheses, and suggest means to test them, to facilitate future research into host movement and microbial community dynamics.


Introduction
Differences in host habitat often lead to profound differences in the composition of a host's microbiome (Kim et al. 2021, Fackelmann et al. 2023 ).These changes in microbial communities can r esult fr om exposur e to differ ent micr obes (thus incr eased competition between incumbent and invading microbes), differences in the environmental habitat of the host (e.g.changes to temper atur e or light exposure), or differences in the host as a habitat (changes in the host that alter it as a habitat for microbes, e.g.str ess or c hanges in diet).On a local scale, such as movement within the 'home-range' of a host (Wolf et al. 2021 ), there is a notable influence of the environment/habitat on the microbiota alongside host-specific c har acteristics suc h as sex (Corl et al. 2020 ).Ho w e v er, in or ganisms whic h undertake periodic or occasional long-distance movements (e.g.migr ations), ther e is frequently an associated major change in the environment, which may significantly affect the microbiota-thus we focus here on long-distance mov ements r ather than short-distance (i.e.homer ange) mov ements.Giv en that ∼1800 birds (Rolland et al. 2014 ), ∼900 fish (Delgado and Ruzzante 2020 ), and numerous mammals and insects (Fudickar et al. 2021 ) are considered migratory (not to mention the frequent occurrence of long-distance dispersal e v ents; Gillespie et al. 2012 ), there is both a need and opportunity to understand how host-associated microbial communities are shaped by host mo vements .In this perspective piece, we explore the relationship between host movement (on a br oad geogr a phic scale) and microbiota composition/diversity.We then provide an ecological fr ame work, whic h can be used to de v elop testable hypotheses regarding host-microbe relationships during the hostmo vements .
For simplicity, we focus on two types of long-distance movement which we term 'regular' and 'irregular' movement.We use the term regular to describe repeated and often cyclical movements between geogr a phicall y distinct locations (e.g.organisms migr ating between br eeding and feeding gr ounds; Dingle andDrake 2007 , Dingle 2014a ).Irregular dispersal, on the other hand, r efers to mov ement of a host to a new and geogr a phicall y distinct envir onment wher e suc h mov ements often occur stoc hastically and without accompanied physiological pr epar ation (e.g.hydr oc hor ous seed dispersal, or Daphnia on birds; Green et al. 2008 ), though this category also extends to mor e-dir ected onetime movement events (e.g.eels migrating to spawning grounds at the end of their life; Dingle 2014a ,b ).There is broad overlap between these terms and terms such as 'migration' or 'dispersal' (Fryxell et al. 2011 ), but migration and dispersal carry nuances that limit their clarity for our purpose.For example, migratory movements can occur singularly or repeatedly, each with vastly different effects on the microbiota.In such instances, a singular migr ation e v ent may be mor e similar to a stoc hastic long-distance dispersal e v ent (e.g.migr ations for r epr oduction in semelpar ous or ganisms ;Dingle 2014a ,b ).Furthermor e, the term dispersal is often r eserv ed to describe mov ement for the pur poses of r epr oduction (i.e .mo v ement that r esults in gene flow; Bonte and Dahir el ( 2017 ), though see Vickers et al. ( 2023 ) for arguments in favour of nonr epr oductiv e dispersal), but irr egular long-distance mov ements may occur for myriad other r easons.Finall y, 'migr ation' in a stricter sense may be limited to bidirectional, and often seasonal mo vements , while dispersal could be defined as one-way movement a wa y from a source population.Although these definitions are useful, they favour binary states that do not fit well with the known complexities of many organisms and their microbiomeswe ther efor e use 'r egular' and 'irr egular' dispersal to compar e pr ocesses.
Regularl y moving or ganisms comprise almost entir el y animals, while irregular movement is typical of a broad array of plants and spawning animals and can also include adult hosts (e.g.kelp rafts) or e v en entir e marine comm unities (e.g.biofouled flotsam; Oberbeckmann et al. 2016 ).Ho w ever, w e suggest that many commonalities can be found within each category based on shared ecology rather than the taxonomic identity of the host (Moraitou et al. 2022 ).Although ther e ar e cav eats within these gr oups (e.g.r egular dispersers ar e necessaril y activ e dispersers, but can also undertake irregular dispersal), we employ these terms principally because we suggest that the primary factor that shapes microbiota responses to movement is the large changes in habitat resulting from the type of movement.Consequently, the average habitat of a regular disperser is thus drastically different to the average habitat of an irregular disperser-even where such dispersers belong to the same population (Risely et al. 2018, Turjeman et al. 2020 ).In essence, we suggest that it is not host identity or type that is the k e y determinant of mo vement-associated shifts , but it is the type of movement.
Both regular and irregular forms of movement are distinct from localized mov ements wher e, although the micr obiota may be v ariable , the en vir onmental medium thr ough whic h mov ements occur remains more hospitable relative to the migratory movement medium.For example, although the microbiota may be correlated with an organism's 'home range' or environment (Wolf et al. 2021 ), the differences in ecological conditions (e .g. food type , temperatur e, and salinity) ar e small, r elativ e to those found across a migratory journey (Lewis et al. 2017 ).Indeed, migratory movements are, beyond their simple geographic change, also examples of envir onmental c hange (Tøttrup et al. 2008, Eisa guirr e et al. 2019 ).Envir onmental c hanges ar e well-established in sha ping the micr obiota (Rothsc hild et al. 2018, Casto-Rebollo et al. 2023 ), howe v er, the extent of changes within a migratory movement may far exceed the degree of change observed within a singular environment.
Although long-distance movement is commonplace among plants and animals (Gillespie et al. 2012, Rolland et al. 2014, Delgado and Ruzzante 2020, Fudickar et al. 2021 ), w e kno w little about how established microbiota are affected when their hosts undergo long distance mo vements .Yet, these mo vements are somewhat analogous to a perturbation or community disturbance, whic h hav e been fr equentl y linked to c hanges in the micr obiota (Sommer et al. 2017, Santillan et al. 2019 ).For some host organisms, these disturbances principally alter the microbiota through changes in the dominance of core and noncore taxa (Pearman 2023 ), while for others core taxa remain persistent throughout the migr atory pr ocess (Wu et al. 2018, Víquez-R et al. 2021 ).These data, alongside emer ging r esearc h, highlight the pr ofound impact (such as high species turnover/beta-diversity, or substantial restructuring) of long-distance movements on established microbiota (Bierlich et al. 2018, Skeen et al. 2021 ) in response to new envir onments and exposur e to nov el micr obes.Ne v ertheless, current studies largely represent discrete timepoints within continuous processes (because e.g.sampling at stopover sites within a mi-gration is straightforw ar d, while sampling birds in flight is much less so).
Identifying how long-distance movement events impact hostassociated microbial communities will be k e y to understanding fundamental ecological processes in highly dynamic environments and predicting how species on the move may perform in their new en vironments .Recent work by Grond et al. ( 2023 ) revealed that the microbiome plays a significant role during shorebird migration through influencing weight gain by synthesis of fatty acids.Such processes in shorebirds are also observed during summer weight gain in hibernating bears, highlighting the influence of the microbiome in shaping core survival processes of multiple host species (Sommer et al. 2016 ).These data thus support the hypothesis that the migr atory pr ocess is a disturbance, albeit an extreme one, leading to changes in the selective pressures on a comm unity, whic h in turn result in compositional shifts in the micr obiota (Sc hmid et al. 2023 ).Nonetheless , as stated abo ve , with successive migratory events the 'average' habitat becomes closer to the migr atory habitat-whic h may result in movementassociated changes falling within the intermediate disturbance hypothesis.
Changes in the microbiota that may result from irregular longdistance mov ements ar e r ar el y studied, likel y due to the inherent difficulties in sampling or ganisms, whic h disperse stoc hasticall y and acr oss extr emel y v aried distances (Gillespie et al. 2012 ).Howe v er, we suggest that some insights into these processes may be dr awn fr om tr anslocations-whic h ar e (fr om a host perspectiv e) an unpredictable change in habitat (although there is an absence of a migratory environment in the same sense as a migrating organism).Where possible we draw on the scant literature examining irregular dispersal and attempt to fill in the blanks with simulations and translocation based liter atur e (e.g.Webster et al. 2020 ).
Global anthropogenic change is leading to major range shifts for a swathe of or ganisms-r esulting fr om a melting pot of species in vasions (e .g. through competiti ve exclusion; Moone y and Cleland 2001 ), habitat change/destruction (Taylor and Stutchbury 2016 ), and shifting dispersal pathways (e.g.ocean currents) (García Molinos et al. 2022 ).Such range shifts are underpinned by dispersal of hosts, and thus understanding host-microbiome dynamics with regards to irregular dispersal is essential.Indeed, given the already established role of the microbiome in invasion success of a diversity of taxa (e.g.codispersed microbes are not beneficial in legumes, but are in some algae; Simonsen et al. 2017, Bonthond et al. 2021 ) there are practical uses for understanding the microbiota of dispersing or ganisms.Specificall y, understanding the effects that the host dispersal process has on microbiota structure may aid in identifying which taxa may be of greater invasion concern, especially as these invasion events occur with greater frequency.
Independent of host mo vements , microbes continually relocate and disperse within and among hosts and environments (Burns et al. 2017, Stothart et al. 2021, Custer et al. 2022 ).Here, we focus on the effects of long-distance host movements on associated micr obial comm unities and thus, except wher e explicitl y stated, movement and dispersal refers to that of the host rather than of microbes (Dingle and Drake 2007 ).Nevertheless, we acknowledge that micr obe mov ement will be affected by, and inter act with, host movement and associated processes (Bo and Kohl 2021 ).

Challenges and emerging trends
Studying host microbiota during migration often r equir es navigating geo-political barriers and sampling organisms underway (e.g.flying birds).As a result, most studies focus on co-occurring r esident/migr ant individuals (Risel y et al. 2017, 2018, Bierlich et al. 2018, Turjeman et al. 2020 ), or examine start, stopover, and end points of migratory movements (Skeen et al. 2021, Thie et al. 2022, Grond et al. 2023 ).Nevertheless, some patterns appear to be emer ging.In migr atory birds, micr obiota ric hness decr eases immediately following long-distance movement and steadily recovers during residency (Risely et al. 2018, Skeen et al. 2021 ).In other bird species, microbiota diversity has been observed to correlate negativ el y with migr atory distance when sampled at a stopover site (Thie et al. 2022 ); such results are found across a range of migr atory or ganisms (Bierlic h et al. 2018, Skeen et al. 2021 ).When the results of Bierlich et al. ( 2018 ) and Vendl et al. ( 2019Vendl et al. ( , 2020 ) ) ar e consider ed together, an inter esting pictur e emer ges that aligns with patterns observed in birds; the migratory process is associated with a decline in species richness, but recovers following migration.Meanwhile, despite the relative paucity of research on micr obiota of mor e stoc hastic host mov ements, r afts of the brown alga Sargassum in the Northern Hemisphere show higher microbial turnover than coastal or attached hosts (Michotey et al. 2020 ), a pattern mirr or ed in rafts of brown algae in the Southern Hemisphere (Pearman 2023 ).
Thus far, we have focused on the microbiota in a broader sense, but whether such a community is internal or external to the host will also make a difference to the influence of long-distance movements.In an artificial movement setting (translocations), equine caecal microbiota declined in species richness during tr av el (Perry et al. 2018 ), whereas the bovine nasopharyngeal microbiota incr eased in ric hness (Chai et al. 2022 ).Suc h shifts in microbiota structure may result from differences in exposure levels during mov ement, prior exposur e to mov ement envir onments, or differential effects of mov ement-str essors on the microbiota (e.g.hosts adapted to migration versus those not adapted).
The frequency and nature of host movements can alter host exposur e le v els and metabolic r equir ements, and impact hostmediated processes on the community.Distinctions between the impacts of regular and irregular movements on the microbiota may arise due to the evolutionary significance and adaptation of man y or ganisms to migr atory mov ements (e.g.imm une function is affected by migration; Eikenaar et al. 2020 ; and immunogenetic variation is in turn linked to micr obiota structur e; Gr een et al. 2008 ).Meanwhile irr egularl y moving or ganisms may lac k suc h ada ptations (or e v en be aided by losing their original micr obiota; Simonsen et al. 2017 ).In essence, there are differences in the microbiota of organisms, which migrate repeatedly throughout their life and those of or ganisms whic h onl y migr ate once.If long-distance mov ements ar e fr equent or r epeated, their r egularity means that hosts could de v elop ada ptations to long-distance movement (Dingle andDrake 2007 , Turjeman et al. 2020 ), which may buffer the microbiota from external stressors .T hese regular host movements are fairly predictable and thus straightforw ar d to study, and indeed considerable research has been carried out on the microbiota of migratory organisms (e.g; Kreisinger et al. 2015, Le wis et al. 2017, Bierlic h et al. 2018, Gr ond et al. 2018, Risel y et al. 2018, Wu et al. 2018, Vendl et al. 2019, Turjeman et al. 2020, Skeen et al. 2021, Obr oc hta et al. 2022 ).In contrast, the oftenunpr edictable natur e of irr egular host mov ement is r eflected in r elativ el y fe w studies on micr obiota c hanges with suc h e v ents.

Insight from simulations
Although empirical studies are sparse and moving organisms are difficult to study, simulations allow testing and de v elopment of hypotheses.When informed by empirical data for parameterization and ground truthing, such simulations can enhance our understanding of the implications , processes , and patterns involv ed in micr obiota c hanges with host movement.Depending on the scale of interest (and availability of information), models can be parameterized to reflect site-specific conditions or a whole-or ganism av er a ge that aims to balance model complexity, fit, and 'usefulness'.To this end, we employed demonstration models exploring micr obiota-mov ement dynamics, these models a ppr oximate basic comm unity dynamics and assembl y/turnov er ov er sequential time steps/generations and can incor por ate v ariation in envir onmental selection, and its interaction with species traits, and changes in exposure to environmental species pools through time (Box 1 ; https://github.com/wpearman1996/Dispersal _ simulations ).

Bo x 1. Ov ervie w of simulation methods.
Modelling was performed using the ecolottery pac ka ge (Munoz et al. 2018 ) implemented in R Statistical Softwar e (R Cor e Team 2022 ).Models comprised communities of simulated species representing a generic host microbiota.We initialized communities using species lists deriv ed fr om water samples collected either as part of a coastal macroalgal sampling effort (Pearman et al. 2023 ), or off-shore sampling from the Munida Microbial Observatory Time Series dataset (Lockwood et al. 2022 ).Each species was assigned a trait drawn on a univariate scale.Individual fitness in eac h gener ation was calculated based on the distance between traits (i.e. the individuals optimal environment) and the current environment (a location on the same univariate scale that varied to replicate host movement) and the amount of tr ait ov erla p with other individuals (i.e.competition).At each time step all individuals had a chance of dying based on their r elativ e fitness, with r emov ed individuals immediately replaced by individuals of surviving species or immigrants from an environmental species pool.
T he en vironment was defined as the complete set of selectiv e pr essur es imposed on a community from both biotic and abiotic sources, rather than being strictly abiotic-taking the vie w that fr om a micr obial perspectiv e , the en vironment encompasses the host.This a ppr oac h means that the selection imposed on community is univariate and analogous to the first principal component of all selective pressures .T he community of simulated species transitioned from their 'origin' environment, through a 'movement' environment, and into a 'destination' envir onment.Eac h envir onment (origin, mov ement, or destination) exerted selection on this scale via a Gaussian function centr ed ar ound the selection par ameter of eac h envir onment with a sigma value of 0.05.Assuming r elativ e similarity between origin and destination environments (e.g.coastal regions separated by open ocean), selection was centred around 0.3 for the origin and 0.4 for the destination.Meanwhile, selection in the mo vement en vironment (e .g. the open ocean itself) was centred around 0.7 to reflect that movement often occurs acr oss envir onments that ar e dissimilar to either the origin or destination en vironments .To em ulate natur al envir onmental gr adients, tr ansitions fr om one environment to the next were gr adual.Ther efor e, selection on the first 10% and last 10% of generations in the Movement period linearly increased or decr eased to/fr om 0.7 to smooth the tr ansition fr om/to origin or destination conditions.For a detailed description of simulation methods , including code , see the Supplementary Material .
We br oadl y defined the micr obiota as an y delineated micr obial comm unity associated with a host, with membership being any microbe with a nonzero abundance .T his definition thus includes transient microbes within the community, though man y suc h micr obes occur at suc h low abundances that they are lost rapidly or have no significant effect on the community ( Supplementary Figures ).Although we did not define a core micr obiome acr oss all sim ulations (as eac h host is modelled independentl y r ather than within a host population context), we divided the community into abundant/rare taxa based on an av er a ge abundance acr oss the sim ulations (abundant taxa were those with an average abundance > 0.1%).Host organisms comprise a br oad arr ay of microbiota (e.g.skin, gastrointestinal, and or al), eac h exposed to differing envir onments and selection pr essur es .T her efor e, to understand the effect of differing exposur e le v els (i.e.understanding internal/external communities), we varied the species richness of the environmental community.From our simulations, we then analysed species richness, dissimilarity (Bray-Curtis), and functional diversity (based on micr obial tr ait v alues).
Despite their broad parameterization, our simulations produce results that are analogous to those observed in real plant and animal communities (e.g.Bierlich et al. 2018, Risely et al. 2018, Webster et al. 2020, Skeen et al. 2021, Chai et al. 2022 ), suggesting that e v en simple models can help explain ecological processes.Our simulations demonstrated that long-distance movement influences the membership and r elativ e abundance of host microbiota (Fig. 1 A and B; Supplementary Figs 2 -4 ).Micr obiota, whic h underwent irr egular mov ement experienced a shift to w ar ds a community dominated by microbes with optimal fitness in the mo vement en vironment (Fig. 1 C).Conversely, microbiota undergoing r egular mov ement, suc h as during cyclical migr ations, wer e dominated by microbes with an intermediate fitness across all mov ement sta ges (Fig. 1 D), r ather than optimal fitness during one stage and subpar fitness in all other sta ges, r esulting in r elativ el y consistent patterns in richness and comm unity structur e. Host mov ement could, ther efor e alter micr obiome functional div ersity or structure in such cases, as has been observed in migratory birds (Grond et al. 2023 ).Irregular movements also resulted in a compositional shift in comm unity structur e a wa y from the initial community (Fig. 1 E), while for regular movements we found that, following the acclimation phase, dissimilarity to the initial community was dependent on whether a community was in the origin or destination location (Fig. 1 F).This finding perhaps results fr om r eac hing a point of maxim um dissimilarity (after accounting for cosmopolitan microbes), and transient changes in microbiota structure in the equilibrated phase result from loss/acquisition of microbes in the origin community, as our simulations do not consider long-term changes in diversity of the environmental pools.
Simulations under a scenario of regular migration resulted in an initial increase in species richness (Fig. 1 B), after which richness tended to exhibit a marginal decline during the migratory period-following patterns anticipated under the intermediate disturbance (Connell 1978 ) or intermediate stochasticity hypotheses (Santillan et al. 2019 ).These simulated data, alongside observ ations fr om the skin micr obiota of humpbac k whales (Bierlic h et al. 2018 ) and faecal microbiota of Kirtland's warblers (Skeen et al. 2021 ), paint an intriguing picture of the effects of migration on micr obiota structur e in r egularl y migr ating or ganisms.A migr atory organism's first few migrations can be considered analogous to irregular mo vement events , and can be grouped together as an 'acclimation' phase (Fig. 1 D), wher e comm unities ar e becoming acclimated to movement events .T he repeating nature of regular movement then leads to an equilibration of the microbiome wher e fr equent mov ement is essentiall y the 'ne w normal' and the turnover or change in the community over time has begun to stabilize, resembling the constant low levels of turnover found in pr e vious studies (Mo y a and Ferrer 2016 , Priy a and Blekhman 2019 ).Although such stabilization may be expected throughout the de v elopment of e v en nonmigr ating or ganisms (Gr ond et al. 2018, Rosw all et al. 2021 ), w e suggest that the migration-induced disruption to micr obial comm unities is a k e y determinant of the assembly of a migration-acclimated microbiota found in/on regularly moving hosts.
Microbiota of hosts that move irregularly did not exhibit the same stabilization (Fig. 1 A).T hus mo v ement, if it occurs onl y singularly or for the first instance (e.g.natal dispersal in salmon, or juv enile migr atory birds), is compar able to an instance of irr egular movement with regards to microbiota structure even if such an or ganism ultimatel y will under go r egular mov ement (tr ansitioning to the equilibrium phase of regular mo vement).T he phases of migr ation-micr obiota equilibr ation (acclimation/equilibr ation; Fig. 1 D) and changes in diversity in response to migration are also lar gel y sha ped by system specific factors (e.g.internal/external, or migr ating/nonmigr ating).For example, migr atory individuals of the same host species, cohabiting the same location, can have distinct microbiota to those of nonmigratory counterparts (Risely et al. 2018, Turjeman et al. 2020 ), and in a similar way an internal microbiota may be affected differently by migration than an external microbiota.
Incr eases in ric hness observ ed during the acclimation phase ar e pr oportional to the richness of the environmental pool to which the microbiome is exposed (i.e. the more potential colonizers, the greater the increase in richness) (Fig. 1 A and B).This relationship may explain why travel leads to decreases in richness in gut microbiota (Perry et al. 2018, Obr oc hta et al. 2022 ), while nasopharyngeal microbiota increase in richness (Chai et al. 2022 ).Ther e ar e two principle r easons why exposur e may affect the microbiota of migrating organisms.First, internal communities are perhaps more insulated against extreme environmental fluctuations (e.g.variations in light, temperature, or UV exposure) and thus internal cavities may r epr esent a more stable environment r elativ e to the envir onment experienced by a skin comm unity (Costello et al. 2009 ).Secondly, internal cavities are under reduced colonization or invasion pressure relative to a skin microbiota; potential colonizers of internal cavities must be introduced thr ough br eathing or food consumption compared to a skin micr obiota, whic h is exposed to the entire 'microbial soup' of the envir onment (Gr önr oos et al. 2019 , Harrison et al. 2019, Vila et al. 2019, Ber ggr en et al. 2023, Santos et al. 2023 ).Internal cavities of migr atory or ganisms may be under e v en less colonization pr essur e during migr ation due to fasting or altered feeding regimes, whic h may r educe exposur e to ne w or ganisms during the migr atory process (Klinner et al. 2020 ).
A comparison of skin and faecal microbiota of wild salmon preand post-translocation into a natural environment also supports the hypothesis that exposure can mediate microbiota diversity (Webster et al. 2020 ).These differ ences in exposur e may explain why internal microbiota frequently decline in richness, while external communities increase in richness (Perry et al. 2018, Webster et al. 2020, Chai et al. 2022 ), during movement.Such differences may be amplified during migr atory e v ents wher e c hanging selective pressures as a result of physiological changes may compound with r educed micr obial intake (due to altered feeding r egimes, or fasting; Nic hols et al. 2022 ), leading to an ov er all decline in richness.Although we suggest that physiological changes primarily co-occur with regular dispersal, they can also occur in instances of irregular movements (e.g.re producti ve migration of

The importance of repeated sampling
Our sim ulations pr ovide continuous data acr oss all gener ations of the modelled microbiota (Fig. 1 ), something which is rarely available when collecting data empirically (Jones et al. 2023, Fountain-Jones et al. 2024 ).Ne v ertheless, r epeated measur es of the same individual can provide invaluable insights into drivers of microbial comm unity assembl y (e .g. en vironment, host genetics , and so on; Grieneisen et al. 2023 ).Repeated snapshot sampling of migrating or ganisms has alr eady pr ovided cor e insights into community assembly rules (Skeen et al. 2021 ), and sampling of meerkats (Risely et al. 2022 ) and primates within their home-ranges (Björk et al. 2022 ) has highlighted the v ariable natur e of the microbiota over time, as well the importance of individual identity.Quantifying this on a broader scale and with increasing frequency in migrating organisms would provide an ideal means to quantify the role of environmental distances in shaping community structure relative to the role of the host.Although such questions are of interest, a fr ame work for de v eloping hypotheses ar ound micr obiotamigr ation inter actions is pr esentl y absent.

A fr ame work for micr obiome inter actions with long-distance movement
A conceptual fr ame work could be beneficial for focussing futur e effort, addressing existing shortfalls, and enhancing understanding of microbiota movement dynamics and contrasting patterns of r egularl y and irr egularl y moving or ganisms.Suc h a fr amework enables de v elopment of testable hypotheses and questions and allow identification of commonalities across a range of complex systems.We suggest that microbiota diversity (richness and turnover) and function in dispersing organisms are best explained by three groups of interacting factors: movement frequency (e.g.r egular/irr egular; Fig. 2 ), exposur e (both exposur e to the envir onment, and exposure to other microbes-e.g.internal and internal communities), and selection (imposed by type of movement and associated physiological stress).

Dispersal frequency
We expect frequent long-distance movement to lead to a mov ement-equilibr ated comm unity as the micr obiota r espond to less-stable en vironments , and thus the response to mo vement decreases in magnitude with increasing frequency of movement.In essence , mo v ement thr ough m ultiple envir onments pr omotes se-F igure 2. F r ame work for the study, explanation, and prediction of movement-driven changes to microbiota structure.We propose that postmovement comm unities ar e most influenced by the form of mov ement (i.e.r egular v ersus irr egular) while exposur e to ne w micr obes in the mov ement environment and selective processes, such as host homeostasis, which may promote resistance to change or host decomposition which may acceler ate c hange, inter act to add to, and/or modulate mov ement fr equency effects.lection for micr obes, whic h can persist thr ough all envir onments, and thus repeated movements will stabilize on a 'core' community in the 'equilibrium phase' with reduced turnover relative to 'acclimation' phase .T he corollary of this hypothesis is that hosts, whic h r ar el y under go long-distance mov ement will hav e extr emel y mov ement-affected micr obiota, whic h could explain why the microbiota of some migrating whales and bir ds sho w relativ el y little c hange in comm unity structur e when compar ed to irr egularl y dispersing organisms (Bierlich et al. 2018, Skeen et al. 2021 ).

Exposure
As discussed pr e viousl y, exposur e to external micr obes str ongl y influences changes in community richness (Fig. 1 ).Our simulations suggested that the primary determinant of microbiota movement responses in the equilibrium phase was the richness of the environmental community (Fig. 1 ), because gr eater exposur e to the envir onmental pool of micr obes increases the potential for colonization.While not incor por ated into our simulations, colonization order and priority effects also promote persistence of incumbent microbes (Sprockett et al. 2018 ), which may be another reason why many microbiota typicall y hav e higher ric hness following mov ement.We expect the same principles to a ppl y acr oss tissue types and compartments , with mo v ement-driv en micr obiota differ ences arising fr om differ ent contributions of exposur e and selection acr oss tissues/compartments .For example , internal compartments ma y hav e v ery low envir onmental exposur e, r esulting in them be-ing shielded from both the environment and potential colonizers (Brown et al. 2020, Zhang et al. 2022 ) and may experience relativ el y little mov ement-pr ompted comm unity c hange compar ed to skin micr obiota, whic h hav e m uc h higher exposur e (Webster et al. 2020 ).

Selection
Selection is another determinant of changes in microbiota richness in response to movement, with stronger selection leading to smaller increases in richness ( Supplementary Fig. 2 ).Changes in the microbiota as a result of frequent long-distance movements ar e under pinned by selection, as fr equent mov ement will select for communities that can withstand that movement (i.e. a 'core' micr obiome; Risel y 2020 ).Ho w e v er, r egardless of mov ement fr equency, selection will shape the microbiota in some way, such as when a host activ el y r egulates their micr obiota, potentiall y pr omoting resistance to change (Arnault et al. 2022 ) or alternativ el y, when suc h r egulation is disrupted leading to decline in the selectiv e pr ocesses (Naylor et al. 2017 ).Fasting and fat loading are both fr equentl y observ ed behaviours associated with long-distance migrations, and these in turn impose ad ditional selecti v e pr essur es on the microbiota (Risely et al. 2018, Thie et al. 2022, Grond et al. 2023 ).Conv ersel y, decomposition and deterioration affects microbiota structur e (Pr eiswerk et al. 2018 ), and thus hosts whic h under go degr adation during long-distance mov ements may experience selection to w ar ds degrading/decomposing microbes, leading to greater than expected changes in microbiota structure (Pearman 2023 ).
Mov ement e v ents also lead to the exposur e of a host to a br oad r ange of envir onments (Tøttrup et al. 2008 ), eac h of whic h may r epr esent both a source of ne w micr obes and a source of selection (Thie et al. 2022 ).In turn, mo vements ma y expose hosts to a new suite of diseases and pathogens, which can amplify changes in the microbiome (Zhang et al. 2021 ).Although the relationship between long-distance movements and disease has not been compr ehensiv el y explor ed, some studies hav e pr oposed the use of micr obiome surv eillance as a tool to detect pathogens such as Salmonella (Choi et al. 2021 ).Indeed, given the role of bird migr ation in tr ansmission of avian influenza (Sun et al. 2022 ) and the r ecentl y determined importance of stopov er sites in the establishment of avian pathogens in migrating birds (Włodarczyk et al. 2024 ), there is a clear need to understand the relationship between microbiome composition, pathogens, and long-distance mo vements .
Changes in selection can arise from a variety of sources, and the magnitude of the changes in the microbiota likely arise from the magnitude of selection.One principal source of selection may be the environmental dissimilarity between the origin and mo vement en vironments .For example , flamingos lar gel y r eside in hypersaline en vironments , but migrate through hypo-or meso-saline environments (Amat et al. 2005 ).Suc h extr eme differ ences likel y contribute to changes in the community in the same way that environmental exposure can shape the microbiota.A 'migr ation-equilibr ated' micr obiota could also feasibly occur via means of dormancy, where microbes which are ill-suited to the migratory medium enter a state of a dormancy until exposed to mor e favour able conditions in the origin/destination envir onments, thus tempor aril y bypassing selection that may be acting on nondormant members of the community.For example, the persistence and presence of halophilic microbes in migrating flamingos (Yim et al. 2015, Kemp et al. 2018, Gillingham et al. 2019 ) may be aided b y dormanc y.Although dormanc y has been fr equentl y described in nonhost-associated microbiota (Jones and Lennon 2010, Locey et al. 2020, Sorensen and Shade 2020 ), to our knowledge no studies have yet explored dormancy in microbiota associated with a host.

Consider a tions and application
While our fr ame work (Fig. 2 ) is primarily relevant to preestablished comm unities, r educed v ertical tr ansmission (acquisition of microbes from parents) may be advantageous in fluctuating environments such as those resulting from and/or experienced during, long-distance movement (Bruijning et al. 2022 ).We suggest that while v erticall y acquir ed micr obes ma y ha ve an initial adv anta ge ov er horizontall y acquir ed micr obes (micr obes acquir ed fr om the envir onment), this can be explained by either priority effects (Spr oc kett et al. 2018 ) or selection (Zeng et al. 2017 ).In the case of the former, there is no reason to expect that their presence due to vertical transmission has any bearing on their r esponse to mov ement.Ho w e v er, in the case of vertical transmission of positiv el y host-selected microbes, we would anticipate that these microbes may be more resistant to movement-driven changes and thus under positive selection.These beneficial microbes could be considered holobiotic and may be a source of inertia for resisting movement-prompted changes to the microbiota.Ne v ertheless, disentangling the contributions of different ecological factors to observed microbiota diversity remains critical, especiall y giv en the pr esence and possible inter actions between factors such as priority effects, migration timing, environmental conditions and diversity, and seasonal variability (Zhang et al. 2021 ).
Although microbial communities are highly dynamic, and have been observed to undergo daily (Zarrinpar et al. 2014, Zhang et al. 2023 ) or seasonal cycles (Huang et al. 2022, Schmid et al. 2023 ), ther e r emains str ong e vidence for priority effects on microbial communities (Martínez et al. 2018 , Leopold andBusb y 2020 ).Ho we v er, the influence of priority effects may be linked to both the degree of disturbance induced by long distance movements (e.g. higher disturbance le v els or higher envir onmental v ariability may ov ercome priority effects; Tuc ker and Fukami 2014 ), and the time between movement events.In the latter instance, increased micr obial gener ations between mov ements is likel y to be associated with reduced influence of historic mo vements .All of these nuances are capable of occurring in a 'neutral' host system (where the host is neutral, but selection operates within the community), ho w e v er in systems where the host may actively regulate or maintain a mov ement-ada pted micr obiota suc h priority effects may be of ada ptiv e benefit to the host (Debr ay et al. 2023 ).
Our fr ame work can be applied to various host-associated microbiota, noting that different host taxa may emphasize different elements .For example , seed micr obiota ar e likel y to be mor e influenced by environmental conditions (Morales Moreira et al. 2021 ), while some migratory birds ma y ha v e gr eater host-dir ected mediation of the microbiota (e.g.microbiota-immune system interactions; Bolnick et al. 2014, Leclaire et al. 2019, Fleischer et al. 2022 ).
Ther efor e, although plants are not migratory in the same sense that many birds are, they still undergo long-distance dispersal and can thus still fit within our fr ame work.Specificall y, plants would fr equentl y fall into the acclimation phase (Fig. 2 ) as the adapted phase is found in those or ganisms, whic h under go r epeated mov ements.
Despite the clear and well-documented influence of host processes (Blekhman et al. 2015, Xiong et al. 2021 ), there remains a large body of work which points to microbiota being explained by pr edominantl y ecological processes that are independent of the host (i.e. are host agnostic) (Adair and Douglas 2017, Rothschild et al. 2018, Mukherjee et al. 2021, Moraitou et al. 2022 ).Such hostindependent factors are well-represented in the proposed framework (Fig. 2 ) and, coupled with the knowledge that host-microbe interactions can modify community structure further, this provides a platform from which hypotheses for specific organisms can be de v eloped.

Key hypotheses and how to test them
Using our fr ame w ork (Fig. 2 ), w e pr opose a r ange of testable hypotheses that could giv e impr ov ed insight into microbiotamo vement dynamics , and into the mec hanisms thr ough whic h host micr obiota c hange with host mov ements (H1-H7).We then outline existing and emerging experimental approaches (A-C) that could be used to test these hypotheses.While we focus here on mo vement dynamics , long-distance mo v ements ar e, by their nature, examples of extensive environmental change.

General hypotheses
H1) Strong host selection (e.g.homeostatic regulation) will lead to less microbiota change as a result of a movement event.
Conv ersel y, a lac k of host r egulation will amplify the effect of movement on microbiota structure.H2) The ecological processes gov erning micr obiota turnov er will shift to w ar ds selection [both host-and nonhost imposed; assessed following methods de v eloped by Stegen et al. ( 2013 ) and Ning et al. 2020 )] following long-distance mo vements , but the influence of selection will decline with successiv e mov ements.This expectation is because sudden envir onmental c hanges impose a major envir onmental filter on a microbiota, but successive movements bring this movement closer to the 'av er a ge' envir onment.H3) Due to priority effects (Spr oc kett et al. 2018, Sieber et al. 2019 ), incumbent microbes will be retained throughout the movement period, albeit with decreased abundance, e v en if they are ill-suited to movement.For example, micr obes whic h ar e ada pted to the origin may be r etained thr ough the migr atory pr ocess, e v en if they hav e lo w er fitness during the migr atory pr ocess.Suc h a scenario may arise jointl y fr om priority effects or from dormancy of those microbes during long-distance mo vements .H4) We hypothesize that internal microbial communities will r espond differ entl y (but not necessaril y less dr amaticall y) to migration than external microbiota owing to differences in exposure to environmental microbes.H5) During long-distance mo vements , turno ver of bacteria is ele v ated (driv en by loss of tr ansient/noncor e micr obes) and r elativ e abundances might differ, while the composition of core community remains stable.

Hypotheses specific to regular movement
H6) Hosts undertaking their first migration will experience r a pid uptake of environmental microbes and saturation of micr obial ric hness, and will initiall y show similar patterns in microbiota structure to irregularly dispersing hosts.

Hypothesis specific to irregular movement
H7) Micr obial ric hness will r a pidl y satur ate during mov ement, but micr obial turnov er will continue to incr ease until the host settles in a new environment.
Across the board, longitudinal sampling of host microbiota over the course of regular and irregular movement events could substantiall y adv ance our understanding of how host mov ement can sha pe micr obiota.Prioritizing r epeated sampling of micr obiota, to enable study of microbiota successional trajectories and turnover within changing en vironments , would be a valuable component of such studies.Ho w ever, recognizing that directly following and sampling dispersing hosts is often infeasible, here we suggest a trifecta of empirical means to test the above hypotheses.
A) Manipulative experiments : laboratory mesocosms would enable dir ect contr ol and modification of mov ement/disturbance fr equency, envir onmental exposur e, and selection.Such systems would enable experimental testing of H1-4 and cyclical variation of the controlled environment would enable assessment of the transition from equilibrium to acclimation phases.In essence, these mesocosms would manipulate the environment around the host to mimic the migr atory envir onment (i.e. the environment mov es ar ound the host, r ather than host moving through the environment).Mesocosms with uniquely barcoded, but otherwise identical, bacteria would pr ov e especiall y helpful for testing the effects of selection and immigration in isolation, following those experiments conducted by Cira et al. ( 2018 ).B) Field observations : we suggest that new methods for tracking and opportunistic sampling of dispersing hosts should be explored, with the integration of complementary methodologies an area of particular pr omise.For irr egular mov e-ment, opportunistic sampling combined with movement modelling (e.g.particle drift modelling) (Michotey et al. 2020 ) and genetic inference of origin populations could provide pseudolongitudinal data to test H1, H2, H4, and H5.C) Host proxies : sampling pr oxy 'hosts', suc h as biofilms from marine debris (Oberbeckmann et al. 2016 ) or vessel hulls, could also be informative.Comparisons of v essels, whic h undertake frequent and consistent long-distance trips (e.g.commercial fishing boats or cargo ships) to those which undertake intermittent trips (e.g.leisure boats) would enable the assessment of H2-H3 and H5-H7.Such systems would be particularly useful for understanding how the microbiota of simplistic hosts can be shaped by the movement pr ocess-pr oviding cor e insights on how communities shift in response to movement in the complete absence of host regulation.For example, host regulation of the microbiota in such an instance would be simplified down a basic factor of antibiofouling paints-enabling m uc h mor e dir ect testing.

Outlook
Ultimatel y, gr eater understanding of micr obiota-mov ement dynamics r equir es the a pplication of a r ange of both experimental and observ ational a ppr oac hes .We suggest that mo v ement fr equency, the le v el of exposur e to envir onmental micr obes, and selection (both host-and envir onmentall y imposed) ar e the primary drivers that shape how microbiota respond to long-distance mo vements .T he need for such a framework was prompted by a recognition of the importance of, but relative paucity of literatur e examining, micr obiota r esponse to irr egular mov ement.Our fr ame w ork allo ws us to derive a range of testable hypotheses regarding microbiota dynamics during host-mo vements , which we hope will initiate and guide future research directions.Special focus should be given to disentangling the r elativ e contributions of movement frequency , environmental diversity , and selection in shaping the microbiota, and using these to bridge experimental and observational studies .T hese studies have critical relevance to biosecurity in terms of disease movement and predicting future microbial scenarios in the context of rapid environmental and climatic change.

Figure 1 .
Figure 1.Influence of movement on simulated microbiota species richness, functional diversity (as traits on a univariate scale), and dissimilarity.Grey shading indicates movement periods.Horizontal axes r epr esent time in generations of simulated microbes.Left-hand panels (A, C, and E) show irregular host movement, while right-hand panels (B, D, and F) show regular host movement.Line colour reflects the relative richness of the environmental pool of microbes during movement (blue = low diversity pool and red = high diversity pool).