Host and nonhost bacteria support bacteriophage dissemination along mycelia and abiotic dispersal networks

Abstract Bacteriophages play a crucial role in shaping bacterial communities, yet the mechanisms by which nonmotile bacteriophages interact with their hosts remain poorly understood. This knowledge gap is especially pronounced in structured environments like soil, where spatial constraints and air-filled zones hinder aqueous diffusion. In soil, hyphae of filamentous microorganisms form a network of ‘fungal highways’ (FHs) that facilitate the dispersal of other microorganisms. We propose that FHs also promote bacteriophage dissemination. Viral particles can diffuse in liquid films surrounding hyphae or be transported by infectable (host) or uninfectable (nonhost) bacterial carriers coexisting on FH networks. To test this, two bacteriophages that infect Pseudomonas putida DSM291 (host) but not KT2440 (nonhost) were used. In the absence of carriers, bacteriophages showed limited diffusion on 3D-printed abiotic networks, but diffusion was significantly improved in Pythium ultimum-formed FHs when the number of connecting hyphae exceeded 20. Transport by both host and nonhost carriers enhanced bacteriophage dissemination. Host carriers were five times more effective in transporting bacteriophages, particularly in FHs with over 30 connecting hyphae. This study enhances our understanding of bacteriophage dissemination in nonsaturated environments like soils, highlighting the importance of biotic networks and bacterial hosts in facilitating this process.


Introduction
Bacteriopha ges ar e viruses able to infect and replicate within bacteria (Rohwer et al. 2009 ).These viruses affect the competition dynamics , structure , and evolution of bacterial populations (Weitz andWilhelm 2012 , Middelboe andBrussaard 2017 ).The coevolution between bacteriophages and their bacterial hosts is considered an important driver for the vast phenotypic and genotypic diversity found in bacterial populations (Koskella and Br oc khurst 2014 ).Mor eov er, bacteriopha ges ar e known to alter competition between bacterial strains (Koskella et al. 2012, You et al. 2022 ), pr omote biodiv ersity (Buc kling and Rainey 2002 ), and mediate horizontal gene tr ansfer (Canc haya et al. 2003, Paul 2008 ).Given the abundance of bacteriophages in nature and their potential impact on the selection and evolution of their host, bacteriophage r esearc h has gained attention in microbial ecology (Koskella and Br oc khurst 2014 ).One of the most significant bottlenecks to understanding the importance of bacteriophages on the evolutionary dynamics of bacterial populations is the mechanisms governing the encounter of bacteriophages and their host.Unlike many of their host bacteria that expr ess activ e mobility, bacteriopha ges are thought to encounter suitable hosts mainly by diffusion (Diaz-Munoz and Koskella 2014, Simmons et al. 2018 ).The typical diffusion constant of a viral particle has been reported to be in the order of ca.0.5-1.0 × 10 −7 cm 2 s −1 , suggesting that the diffusion of bacteriophages b y Bro wnian motion at the nanoscale is limited (Dennehy 2014 ).This limitation is particularly relevant in environments such as soils, where waterborne diffusion is already highly r estricted.In soils, vir al comm unities hav e been shown to exhibit spatial structuring, thus hinting at the role of dispersal limitation for community assembly (Santos-Medellin et al. 2022 ).
Dispersal limitation in soils r esults fr om the high density of porous spaces with varying degrees of water saturation, the diversity of solid surfaces with varied physicochemical properties, and the intermixing of solid-liquid-gaseous phases (Crawford et al. 2005, Or et al. 2007 ).Water saturation is a determining factor in the composition of bacterial communities, as it affects the diffusion and bioavailability of nutrients (Crawford et al. 2005 ), as well as the dispersal of bacteria themselves (Kohlmeier et al. 2005, Or et al. 2007 ).Mor eov er, the soils' physical structure (i.e.connectivity, the distribution of water and nutrients) determines the access to resour ces b y coexisting or competing populations of micr oor ganisms (Erktan et al. 2020, Dubey et al. 2021 ).Variation in the connectivity of the soil pores and the water films results in the creation of refuges impacting the probability that consumers/predators encounter food sources/hosts (Erktan et al. 2020 ).In this context, the hyphae of filamentous fungi provide indispensable opportunities for encounters and interactions between other members of the soil microbiota (Bielcik et al. 2019 ).The mycelium of fungi and oomycetes have been shown to effectiv el y connect air-filled water-unsaturated soil pores (Kohlmeier et al. 2005, Wick et al. 2007 ).The de v elopment of continuous liquid films along hyphae provides a network for the dispersal of bacteria, called fungal highways (FHs) (Kohlmeier et al. 2005 ).
The biotic networks created by FHs could contribute to or hinder the movement of bacteriophages from host-poor to host-rich ar eas in fr a gmented soil habitats .For instance , pr e vious studies indicated that bacteriophage diffusion along mycelia is limited in unsaturated environments (You et al. 2022 ).Under watersaturated conditions, mycelia have been shown to promote bacteriopha ge r etention depending on the physicoc hemical pr operties of the bacteriophage and the mycelium (Ghanem et al. 2019 ).Ne v ertheless, tr ansport of bacteriophages along FHs has been observed with the help of hyphal-riding bacteria acting as carriers (You et al. 2022 ).You et al. observ ed that Esc heric hia T4 virus r e v ersibl y attac hes to a motile nonhost bacterium ( Pseudomonas putida KT2440), and uses it as a carrier to ride across hyphae of the oomycete Pythium ultimum to r eac h its host (i.e.Esc heric hia coli ).In contrast, in the absence of carrier bacteria, the bacteriophage was not detected across the unsaturated area connected by the hyphae (You et al. 2022 ).This facilitated transport of bacteriophages by riding on nonhost bacteria was also shown to take place in biofilms (Yu et al. 2021 ) and in soils (You et al. 2022 ).Together, these studies highlight the importance of bacterial carriers on bacteriopha ge dissemination, r esulting fr om the combined effect of diffusion and transport in the absence of r epr oduction in a nonhost carrier.In contrast, unassisted dissemination of bacteriophages on FHs by diffusion has been shown to be negligible.
While prior studies used nonhost carriers to demonstrate bacteriopha ge tr ansport along FHs, the impact of host carriers and the possibility of bacteriophage replication within the host during dissemination has not yet been investigated.Our study bridges this knowledge gap by addressing the following questions: (1) to whic h degr ee do bacteriopha ges diffuse unassisted within liquid films established on the surface of abiotic or biotic (e.g.FHs) dispersal networks?(2) What is the impact of host bacterial carriers compared with nonhost carriers on the rate of bacteriophage dissemination along biotic and abiotic networks?This study demonstrates the combined positive effect of transport and multiplication in the hosts carrier for the dissemination of bacteriophages in structured environments connected via biotic networks.

Microorganisms and culture conditions
The experiments were conducted with the soil bacterium P. putida (Benedetti et al. 2016 ).Two motile strains of this species, strains DSM291 and KT2440, were used as host and nonhost carrier, respectiv el y.The noncompatibility of the second strain with the two bacteriophages was controlled beforehand via lysate plate assay.Strain DSM291 was purchased from the German Collection of Micr oor ganisms and Cell Cultur es (DSMZ).Bacterial cultur es wer e performed on standard nutrient agar or nutrient broth (NB) pur-c hased fr om Sigma Aldic h.Tw o my oviridae bacteriophages specific to strain DSM291 (DSM100069 and DSM100071, hereafter referred to as P1 and P2), were also purchased from DSMZ.Both bacteriopha ges wer e pr opa gated, purified and counted as described pr e viousl y (Bonilla et al. 2016 ).After pr opa gation the number of plaque forming units (PFU) in the lysates was established by a PFU counting assay with the DSM291 strain as host.Lysates were stored at 4 • C. Incubation temperature was 26 • C. Pythium ultimum , which has been used in pr e vious studies in bacteriophage-FHs dispersal (You et al. 2022 ), was inoculated in M9 mineral liquid medium (Sigma Aldrich) and incubated at room temperature under agitation (Lab Shaker, Adolf Kühner AG) at 120 rpm for 5 da ys .The mycelium was fr a gmented in a 50-ml Falcon tube (Corning) using an ULTRA-TURRAX ® (IKA ® T18 basic) at max speed for 10 s, and then washed three times with physiological water (0.9 g l −1 NaCl).Hyphal fr a gment density was assessed with a Neubauer chamber (BIOSYSTEMS ® 0.01 mm).

Experimental systems
Two experimental devices (called her eafter: 'bacterial tr ail' and 'bacterial bridge'; Figure S1 , Supporting Information ) were used to e v aluate diffusion and activ e tr ansport of bacteriopha ges in model abiotic networks .T he devices were 3D-printed with a heatr esistant hydr ophilic material (3DM-HTR140 resin), and the design, validation, and instructions for their production are described in detail in a pr e vious publication (Kuhn et al. 2022 ).The liquid film in the 'bacterial trail' was created by adding 800 μl of NB in all the wells.In this device, the thickness of the liquid film formed is in the magnitude of 1 mm, which has been shown to result in a hydraulic flow that facilitates diffusion of nonmotile particles, including nonmotile bacterial cells (Kuhn et al. 2022 ).In the second device ('bacterial bridge'), the liquid film was established at the horizontal top of the de vice, wher e two vertical columns allo w ed a connection by capillarity.The 'bacterial bridge' devices wer e pr epar ed by first filling the ca pillaries of the columns with NB (with the device upside down).Then, the device was placed upright betw een tw o w ells in a 24-w ell cell culture plate (Costar ® 24, Corning) filled with 2.75 ml of NB.Finally, in order to establish the liquid film, the top of the capillaries was connected by adding extra medium on the horizontal part, on each side of the central column.The biotic networks wer e cr eated using 'fungal drops' (Buffi et al. 2023 ), which consist of media droplets placed at a fixed distance on a tr anspar ent surface to enable direct visualization ( Figure S1 , Supporting Information ).FHs of P. ultimum were gro wn b y inoculating drops of m ycelial suspension (30 m ycelial particles μl −1 ) into a Petri dish treated for cell culture (100 mm × 20 mm Cell Cultur e Tr eated; Corning).In eac h Petri dish, six pairs of 15 μl drops were deposited at a distance of 0.8 cm, each pair w as separated b y 1.5 cm.The plates were incubated at room temper atur e (25 • C) and with 70% r elativ e humidity to avoid e v a por ation.After 4-5 da ys , the pair ed dr ops wer e connected by P. ultimum mycelia.The connected drops were observed under a stereoscope (NIKON SMZ18) or an inverted microscope (EV OS FL, EV OS M5000, Invitrogen) in order to count the number of connecting hyphae.

Bacteriophage diffusion in an abiotic system
To determine if the bacteriophages diffuse through the liquid film generated on the 'bacterial trail' device, the initial and final PFUs in the inoculation and end wells wer e compar ed after 20 h.The de vices wer e pr epar ed as indicated abov e.After that, 10 μl of a concentr ated bacteriopha ge solution was added to the inoculation well (PFU absolute numbers P1 = 4.2 × 10 6 PFU; P2 = 7.4 × 10 6 PFU).Six independent replicates per bacteriophage strain were performed.The 'bacterial trail' devices were incubated under sterile conditions for 20 h.Then, 150 μl were sampled from the end wells avoiding aspirating liquid from the trail.Bacteriophage concentration was measured as described previously (Bonilla et al. 2016 ), but given the small volume, only the centrifugation and c hlor oform purification steps were performed (no filtration).The lysate plate assays were performed on 6 mm Petri dishes adjusting the v olume accor dingly to half the volume of a r egular assay.Onl y a 1:10 dilution was performed from the purified lysates.PFUs were r ecorded after ov ernight incubation at 26 • C. The absolute number of bacteriophages was estimated by calculating PFUs based on the sample volume (150 μl).

Contribution of mycelia to bacteriophage dissemination
To assess bacteriophage dissemination on biotic networks, the initial and final bacteriopha ge concentr ation was compared between two drops connected by P. ultimum mycelia.One of the drops was inoculated with 10 μl of bacteriophage lysate.After 20 h, the connected dr ops wer e sampled.For eac h r eplicate, 6 dr ops of 15 μl were pooled in 810 μl of physiological water (dilution 1:10).Six inde pendent re plicates were performed for each bacteriophage str ain and eac h hyphal category.Six controls without P. ultimum FHs were performed.The bacteriophage clean-up and lysate plate assay pr ocedur es wer e the same as described abov e.The absolute number of dispersed bacteriophages was estimated by calculating the PFUs based on the sample volume (15 μl).

Impact of host and nonhost carriers in acti v e transport of bacteriophages
The 'bacterial bridge' device was used to determine the effect of activ e tr ansport by motile bacteria (host or nonhost carriers) in an abiotic system.The initial and final bacteriophage concentrations wer e compar ed after inoculating the start well with 100 μl of bacteriopha ge l ysate (1.3 or 6.4 × 10 8 PFU) in the presence of the respective bacterial carrier (DSM291 as host or KT2440 as a nonhost) at an optical density of 0.7-0.8.For each carrier and bacteriophage str ain, six independent r eplicates wer e performed.The de vices were maintained under sterile conditions for 20 h.Then, 1.5 ml was sampled from the end well.The samples were then cleaned up and plated as described before (dilutions 1:1 to 1:10 8 ).The absolute number of tr ansported bacteriopha ges was estimated by calculating the PFUs according to the sampled volume (1.5 ml).In order to assess the effect of carriers in FH, 10 μl of the respective bacterial inoculum was coinoculated into the same drop with the bacteriophages (3.6 × 10 6 PFU).After 20 h, each final drop was sampled (15 μl) and mixed into 1.485 ml of physiological water (1:10 2 dilution), and quantified as indicated abo ve .

Sta tistical anal ysis
To assess whether the biotic network has a significant effect on the dispersal efficiency of bacteriophages, a linear model was generated by including the dissemination efficiency of bacteriopha ges with r espect to the system (abiotic vs biotic).The bacteriopha ge str ain (P1 vs. P2) was added to the model as a covariate since the strain and the interactions between the experimental system and the strain were not significant (in the full model).We performed a Two-Way ANOVA analysis of this adjusted model.Each compar ativ e gr oup comprised 12 v alues.To determine the impact of host and nonhost bacterial carriers on bacteriophage transport, a linear mixed-effect model was created to explain the variation in phage dissemination efficiency.In this model, phage dissemination efficiency was used as the response variable.We included the experimental system (bacterial bridge vs. fungal hyphae), the carrier bacterium (host vs. nonhost), and the bacteriophage strain (P1 vs. P2) as fixed effects.We then performed a Two-Way ANOVA analysis with the full model.Each comparative group comprised 12 values.All analyses were carried out within the R statistical environment (Team 2021 ).

Dissemination of bacteriophages along abiotic and biotic dispersal networks in the absence of carrier bacteria
In order to measure the efficiency of bacteriophage dissemination without multiplication along abiotic and biotic netw orks, w e first e v aluated their diffusion in liquid films on abiotic and biotic (mycelia of P. ultimum ) surfaces.Diffusion along abiotic hydrophilic surfaces was studied in the 3D-printed 'bacterial trail' device; Fig. 1 A).Both bacteriophages, P1 and P2 were found to disseminate through the liquid films genereated with this device (Fig. 1 A).To establish the efficiency of bacteriophage dissemination ( ε), the ratio between the number of viral particles in the inoculation well and the number detected in the end-well was calculated after 20 h of incubation.The efficiency of dissemination of bacteriophages in the 'bacterial trail device' ( ε trail ) was calculated to be 1.47 × 10 −5 for P1 and 4.00 × 10 −6 for P2, meaning that only a handful of bacteriophage particles in a million disseminated through the liquid film.Next, we evaluated the efficiency of dissemination of bacteriophages across the liquid films formed on FHs formed by the mycelia of P. ultimum ( ε FH ).In FHs, the number of hyphae connecting two drops could not be controlled and w as variable betw een re plicates.In ad dition, the exact n umber of hyphae crossing adjacent drops could not be precisely quantified when the drops were connected by more than 30 hyphae (due to m ultiple ov erla pping hyphae).Ther efor e, the number of connecting hyphae was estimated as a range with three categories: drops connected by less than 20 hyphae, connected by 20-30 hyphae, and connected by over 30 hyphae (Fig. 1 B).The two bacteriophage strains disseminated along FHs only when drops were connected by 20-30 or over 30 hyphae .T he ε FH was calculated as the ratio between the number of viral particles in the inoculation drop and the number detected in the connected drop after 20 h.The ε FH increased one order of magnitude with an increasing number of connecting hyphae for both bacteriophages (Fig. 1 C).Even though bacteriophage dissemination on mycelia was still limited, ε FH was higher by one to three orders of magnitude than ε trail (Fig. 1 D).This differ ence was statisticall y significant when comparing the two dispersal networks (abiotic vs. biotic), but not when comparing the two bacteriopha ge str ains ( Table S1 , Supporting Information ).A summary of the dissemination efficiency under different treatments is present in Table S2 ( Supporting Information ).

Impact of host and nonhost bacterial carriers on bacteriophage dissemination along dispersal networks
We next measured the effect of host and nonhost bacteria as carriers .In this case , dissemination includes not onl y the tr ansport, but also the possibility for multiplication of the bacteriophage within the host carrier.To reduce the potential for dissemination in the absence of carriers, the 'bacterial bridge' device was used.In this de vice, activ e swimming has been shown to be r equir ed   S6 ( Supporting Information ).
The statistical significance of comparisons between treatments are shown on the box plots .T he sketch in (A) was modified from Kuhn et al. ( 2022 ).
for bacterial dispersal (Kuhn et al. 2022 ).Accordingly, in the absence of a bacterial carrier, no dissemination of bacteriophages was found.The dissemination efficiency of bacteriophages transported by the two motile bacterial carriers ( ε Bridge ) corresponded to 3.63 and 7.13 × 10 −4 , for the host (strain DSM291) and nonhost (strain KT2440), respectively.T hus , the host bacteria were 5.04 × 10 4 times more efficient for dissemination (combined effect of transport and multiplication) of bacteriophages than the nonhost carrier bacteria (Fig. 2 ).The same experiments were then performed using the mycelia of P. ultimum .The dissemination of bacteriophages in FHs in the presence of a nonhost bacterial carrier ( ε FHcarrier ) did not vary significantly across the experimental systems for the two bacteriophage strains.In contrast, the host carrier enhanced dissemination across the experimental systems .Moreo ver, ε FHcarrier was the most efficient between drops connected by over 30 hyphae.The differences betw een netw orks (abiotic/FHs), carriers (host/nonhost), bacteriopha ge str ains (P1/P2), and the combinations between those factors on the ε of bacteriopha ge wer e all statisticall y significant ( Table S4 , Supporting Information ).A summary of the dissemination efficiency in the presence of host and nonhost carriers under different treatments is present in Table S5 ( Supporting Information ).

Discussion
Our study determined the effect of abiotic and biotic dispersal networks on the dissemination of bacteriophages in the presence and absence of host or nonhost carrier bacteria.The factorial design varying the type of network and bacterial carriers one at a time, enabled us to demonstrate the importance not only of diffusion, but also of transport by bacterial carriers and multiplication within the host, in the dissemination of bacteriophages on biotic dispersal networks formed by mycelia (FHs) in the presence of bacterial host carriers.Dissemination in the absence of carrier bacteria: despite of a v ery low bacteriopha ge dissemination efficiency in the 'bacterial tr ail de vice', we found bacteriopha ge dissemination in both the abiotic and the FHs systems over a distance of 40 mm after 20 h (Fig. 1 ).Diffusion in the 'bacterial trail device' was one to three orders of magnitude lo w er than in FHs.Ho w e v er, bacteriopha ge dissemination in FHs only occurred in the presence of over 20 hyphal connections (Fig. 1 ).Although a direct comparison between the two systems is difficult as the 'bacterial trail' device may provide different connectivity and physicochemical conditions than hyphal surfaces (e.g.due to the secretion of hyphal surfactants), our results suggest that bacteriophage dispersal on hyphal surfaces may take place (Fig. 1 ).Observing bacteriophage dissemination in the presence of high hyphal numbers suggests that bundling neighbouring hyphae may have caused a volume increase of liquid paths, thereby facilitating phage dissemination.It also may explain the difference to results found by You et al. ( 2022 ), who did not find bacteriophage dispersal along P. ultimum hyphae in the absence of bacterial carriers.You et al. ( 2022 ), applied an Escherichia T4 virus to an existing P. ultimum network already bridging an air ga p between a gar patc hes.Although the r easons for the observ ed difference may be multifaceted (e.g.bacteriophage size, bacteriopha ge load, tr av el time , distance , and so on), we propose here that a combination of hyphal abundance and secondary hyphal growth ma y ha ve promoted the dissemination of bacteriophages P1 and P2 in this study.In contrast to static abiotic networks, the activ e gr owth of FHs can influence dissemination.For instance, tips of growing hyphae are considered to be more hydrophilic than older hyphae and, hence, allow for broader liquid films and better diffusive dissemination of bacteriophages (Kohlmeier et al. 2005 ).
Dissemination in the presence of carrier bacteria: for the first time, we quantify the effect of closely related host and nonhost bacterial carriers on bacteriophage dissemination.The transport of bacteriophages by the host and nonhost carriers gr eatl y enhanced the dissemination in both types of systems investigated (Fig. 2 ).Pr e vious studies of bacteriophage dispersal associated with bacterial carriers have used nonhost carriers to assess the impact of transport on the encounter of bacteriophages with their host (Yu et al. 2021, You et al. 2022 ).Another study demonstrated that this mechanism of bacteriophage dispersal is widespread and involv es a lar ge r ange of nonhost bacteria, suggesting that carrier bacteria-facilitated bacteriopha ge tr ansportation can be a highl y r ele v ant pr ocess in soil ecology (You et al. 2022 ).The r ationale for choosing closely related carrier strains is that they are likely to share similar ecological niches and provide similar sets of potential receptors for effective bacteriophage attachment during transport.We found that the dissemination efficiency was indeed higher upon transport by the host, regardless of the network used (Fig. 2 ).
In our experiments with the nonhost carrier, the dissemination efficienc y w as constant regar dless of the netw ork or the number of connections in the biotic network (Fig. 2 ).This suggests that dissemination by transport in the absence of multiplication is likely limited by a fixed variable such as the number of binding sites on the carrier cells or transport by turbulence created by the dispersal of motile bacteria.A r atio abov e 1 in the dissemination efficienc y b y the host signals a positiv e r ole for m ultiplication in addition to tr ansport.The m ultiplication and dissemination by a population of host carriers pr obabl y did not occur sim ultaneousl y (or in a sync hr onized manner in the entire population), as upon burst and release of new viral particles, lysed carriers can no longer contribute to transport, but contribute positively to the number of viral particles that can be transported by other host carriers (or other bacteria in nature) in the dissemination network.Diverse infection strategies in which the bacteriophage does not decimate the population of host cells, including phenomena such as pseudol ysogen y, carrier state, or c hr onic infections (Mantynen et al. 2021 ), could also contribute to w ar ds the positive effect of host carriers on bacteriophage dissemination.Among these mechanisms, pseudol ysogen y does not appear to be likely, as this process results in the injection of the genetic material, but not the active multiplication of viral particles (Ripp andMiller 1997 , Los andWegrzyn 2012 ).In this case, the dissemination efficacy between the nonhost and host carriers would be expected to be equiv alent, whic h was not the case here (Fig. 2 ).The other two pr ocesses ar e mor e likely to contribute to the increase in dissemination efficiency by the host carrier.The establishment of a stable equilibrium in the bacteriophage-host populations with some hosts temporarily resistant to lysis (carrier state; Siringan et al. 2014 ), releasing viral particles without cell l ysis (c hr onic infection; Hoffmann Berling and Maze 1964 ), or simply dispersing before the cell energy is div erted into bacteriopha ge m ultiplication, can all contribute to the increase in dissemination efficiency resulting from transport and multiplication in the host carrier.
Our experimental systems demonstrate both the dissemination of bacteriophages on mycelia and the positive effect of host carriers.Ho w e v er, they also hav e limitations that should be acknowledged.For instance, the effect of hydraulic flow, which has been shown to occur in the r elativ el y thic k (1.35 ± 0.32 mm) liquid films formed on the 'bacterial trail' device (Kuhn et al. 2022 ), might also occur in the mycelial network and contribute to explaining the positive effect of increasing FH connectivity on bacteriophage dissemination.Ho w ever, the exact thickness of the liquid film on mycelium and the existence of hydraulic flow across FHs still need to be measur ed.Like wise, the comparison of bacterial mov ement acr oss the 'bacterial bridge' with FHs in a fixed timeframe does not take into account other factors that might affect bacterial movement in FHs, such as chemotaxis , surfactants , or nutrient pro vision.T hese factors should be considered in future studies , for instance , by using fungal exudates as the medium for bacterial dispersal.Mor eov er, in the futur e, it will be important to de v elop experimental a ppr oac hes to measur e the effect of alternativ e bacteriopha ge infection str ategies (i.e. carrier state or c hr onic infections) to assess the positive effect of the host carrier on bacteriophage transport.
Our results can be considered under a common framework used for other pathogen-host systems in which the dispersal capacity of the parasite is k e y for understanding of its access to the host (Johnson et al. 2019 ).One such framework is the socalled BAM model in which the pathogen niche is determined by biotic factors (B; e .g. host a vailability), abiotic factors (A; e.g.environmental conditions limiting the survival of the pathogen), and the movement capacity (M; e.g.vector-borne) of the pathogen (Escobar and Craft 2016 ).The same elements have been shown here and in previous studies to be r ele v ant for the study of bacteriophage-host dynamics in unsaturated en vironments .T he use of FHs as dispersal networks can be an innov ativ e element to expand the BAM fr ame work, as dispersal networks formed by fungi and fungal-like organisms affect the dynamics of the carriers (host and nonhost) activ el y.In this way, future experiments can investigate disease incidence under different dispersallimited regimes (e.g.heterogeneous distribution of host and nonhost populations connected by biotic/abiotic networks).They will complement pr e vious studies on the r ole of bacteriopha ge tr ansport in the context of inv asiv eness and r ange expansion of nonhost carriers (You et al. 2022 ).We expect suc h futur e work to offer a more comprehensive understanding of the mechanisms of bacteriophage dispersal in soils (Pratama and van Elsas 2018 ) and other spatially structured habitats (e.g.fermented foods, sediments, and surface tissues of plants and animals) in which bacteria, fungi, and bacteriophages actively interact and coexist.

Figure 1 .
Figure 1.Bacteriophage dissemination in liquid films formed on abiotic or biotic networks.(A) Dissemination on the abiotic system was measured using the 'bacterial trail device' (left).The number of bacteriophages propagated to the end well was measured 20 h after inoculation in the start well.Bacteriophage numbers were quantified by PFU.(B) Dissemination on biotic networks was measured on FHs formed by the hyphae of P. ultimum using the drop system after the same incubation time.Drops were connected by different number of hyphae: below 20, between 20 and 30, and above 30 hyphae as illustrated in the images.(C) The number of bacteriophages was quantified before (left columns) and after 20 h for the control (no hyphae), belo w 20, betw een 20 and 30, and abo ve 30 hyphae .(D) Box plots indicating the dissemination efficiency ( ε, calculated as the r atio of final ov er initial number of bacteriopha ges).Onl y the categories in whic h dissemination was measur ed wer e included.All the experiments wer e run in six independent replicates .T he ra w data are provided in TableS3( Supporting Information ).The statistical significance of comparisons between treatments are shown on the box plots .T he sketch in (A) was modified fromKuhn et al. ( 2022 ).

Figure 2 .
Figure 2. Effect of host ( P. putida DSM291) and nonhost ( P. putida KT2440) bacterial carriers on dissemination efficiency.(A) Bacteriophage dissemination efficiency associated to transport in the presence of a bacterial carrier on an abiotic network ('bacterial bridge' device).(B) and (C) Bacteriophage dissemination efficiency in biotic networks formed by P. ultimum FHs.As in the case of Fig. 1 , the number of connecting hyphae was categorized as (B) 20-30 or over 30 (C) hyphae.All the experiments were run in six independent replicates .T he ra w data is provided in TableS6( Supporting Information ).The statistical significance of comparisons between treatments are shown on the box plots .T he sketch in (A) was modified fromKuhn et al. ( 2022 ).