Page 87 · Donor microbiota engraftment after transplantation — strain-level competition in the recipient gut
Fecal microbiota transplant (FMT) is typically performed in patients whose native gut microbiome has already been severely disrupted — most classically by recurrent Clostridioides difficile infection following repeated antibiotic courses. That disruption is not incidental to the therapy; it is part of the mechanism. A depleted community leaves attachment sites, mucin substrates, and metabolic niches unclaimed, and it is that reduced competition — not the donor material alone — that gives incoming strains room to establish.
A healthy gut microbiome exerts "colonization resistance" — a dense, metabolically saturated community that leaves little room or food for newcomers, whether those newcomers are pathogens or would-be beneficial colonizers. Resistance is not a single mechanism but a bundle: competition for shared carbon sources, occupation of physical attachment sites along the mucosal surface, production of bacteriocins and short-chain fatty acids that suppress competitors, and maintenance of a stable pH and oxygen gradient that few outsiders tolerate.
Broad-spectrum antibiotics, recurrent infection, and inflammation collapse this resistance simultaneously — killing susceptible commensals, opening physical space, and leaving behind metabolic byproducts unused. The result is a community with far fewer species holding far more open niche than before.
This is illustrative background, not clinical guidance. The point that matters for the simulation: the emptier the niche, the less resistance a donor strain has to overcome to establish itself.
Not all disrupted microbiomes are equally open. A moderately disrupted community may still retain resilient generalist taxa that persist through antibiotic pressure and continue occupying niche space, resources, and attachment sites. A severely disrupted community — near-total ablation of diversity — leaves substantially more niche genuinely open.
This matters directly for engraftment modeling: donor strains introduced into a severely disrupted recipient face measurably less resident competition than strains introduced into a moderately disrupted one, all else equal. The simulator's "pre-transplant disruption severity" control reflects this — toggling it changes how much resistance the surviving recipient flora offers to incoming donor strains.
It is tempting to picture a disrupted microbiome as an empty container waiting to be filled. It is closer to a partially deflated, still-occupied ecosystem: surviving taxa continue reproducing, mucin layers regenerate, and host immune and dietary signals keep shifting the available substrate landscape. The "open niche" is real, but it is dynamic and contested even before donor material arrives — which is why the strain-level competition in later stages is never a formality.
The donor fecal microbiota preparation — delivered by colonoscopy, enema, nasogastric/nasoduodenal infusion, or oral capsules — introduces a complex, multi-species community into the recipient gut in a single event. What follows in the first hours to days is not "colonization" in a simple sense but a wave of individual strains each independently testing whether the local environment will let them attach and reproduce.
Unlike a single-strain probiotic, an FMT preparation delivers an entire microbial community — hundreds of species and countless strains — into the recipient gut in one dose. This bulk delivery matters biologically: donor taxa arrive already embedded in their own cross-feeding relationships and metabolic networks, which can help some of them establish faster than an isolated strain could on its own.
In the first hours after delivery, donor strains disperse through the lumen and begin encountering open niche essentially at random. Which specific patch of mucosa, which specific attachment site, and which specific competing cell a given donor bacterium meets first is largely a matter of chance and local density — not a deterministic outcome.
This stochastic initial phase is why replicate FMT procedures using the same donor material in different recipients (or even the same recipient at different times) do not produce identical engraftment outcomes. Early colonization sets the stage, but it does not guarantee the final composition.
Initial detection of a donor strain in the days after FMT is not the same as durable engraftment — many strains detected early are later outcompeted or simply fail to persist once resource competition intensifies.
The recipient is not a passive vessel. Residual antibiotic levels, mucus layer thickness, bile acid composition, local pH, immune tone, and diet all influence which donor strains find the environment hospitable. Two recipients receiving material from the same donor can show different initial colonization patterns purely because their post-disruption gut environments differ.
Once donor strains are present in the lumen, the determining question shifts from "did it arrive" to "can it win." Individual donor strains compete with each other and with any surviving recipient strains for the same finite pool of attachment sites and substrates. Strain-level metagenomic tracking — following single-nucleotide variants unique to donor versus recipient lineages — has made it possible to watch this contest resolve in real patients over time.
Two E. coli — one donor-derived, one recipient-derived — are not interchangeable for engraftment purposes even though standard species-level classification would call them "the same." Strain-level competition is fought over subtle differences in metabolic capability, adhesin repertoire, and phage susceptibility that only strain-resolved sequencing can distinguish. This is why modern FMT research increasingly relies on strain-level metagenomics rather than coarser taxonomic profiling — species-level presence/absence can mask which lineage, donor or recipient, actually won.
Classical competitive exclusion theory holds that two populations competing for an identical limiting resource cannot stably coexist indefinitely — one will eventually displace the other. In the gut, this plays out at the level of individual niches: a mucosal attachment site, a specific fiber substrate, a cross-feeding partnership. A donor strain with even a modest fitness advantage in a given niche — faster growth on an available substrate, better adhesin match to the mucus layer — can gradually displace a resident competitor occupying that same niche.
This is a probabilistic, gradual process rather than an instantaneous takeover, and it plays out differently niche by niche across the gut — which is part of why the eventual engrafted community is a mosaic of donor- and recipient-derived strains rather than a uniform donor sweep.
The severity of pre-transplant disruption from Stage 1 directly shapes this contest: fewer surviving, well-adapted recipient strains means fewer competitors a donor strain must displace to claim a niche. This is the mechanistic link between "severe disruption" and "generally higher engraftment success" — it is not that donor strains behave differently, but that they face a weaker field of resident competitors.
This dynamic is illustrative and simplified for the simulation. Real strain-level competition also depends on donor-recipient compatibility, diet, host immune interactions, and factors not modeled here.
Not every donor strain that arrives and competes goes on to establish a stable population. Engraftment success is markedly taxon-dependent: some genera — often robust, oxygen-tolerant, broadly fermentative groups — engraft reliably across many recipients, while others fail to persist even when initially detected. This variability remains incompletely predictable from donor or recipient characteristics alone, and is an active area of ongoing characterization.
Across FMT studies, certain bacterial groups are reported to engraft more consistently than others — broadly, taxa capable of using a wide range of dietary and host-derived substrates, tolerating fluctuating oxygen exposure, and forming close metabolic partnerships with other community members tend to persist more reliably. Narrower specialists, or strains highly dependent on a specific donor-gut context, are reported to engraft less consistently.
The three illustrative "strains" in this simulator (A, B, C) stand in for this spectrum — a robust generalist, a moderate contextual engrafter, and a poor engrafter — rather than representing any single named species.
Even knowing a strain's general taxon-level tendency, whether it engrafts in a specific patient depends on donor-recipient compatibility, the exact composition of surviving recipient flora, diet, medication use, and stochastic early-colonization events described in Stage 2. This combination of many interacting, only partly measured variables is why engraftment prediction for an individual patient remains an active research problem rather than a solved one — taxon-level tendencies inform expectations but do not guarantee individual outcomes.
The percentages shown here are illustrative teaching values for the simulation, not measured clinical engraftment rates for any specific organism.
Because engraftment success varies strain by strain, the resulting post-FMT community is typically a mosaic — some niches held by newly engrafted donor strains, others still held by surviving recipient flora, and some left transiently open pending further competition. This partial, uneven outcome is the expected shape of a successful FMT, not a sign of failure.
A single post-FMT snapshot cannot distinguish a transiently detected donor strain from a durably engrafted one. Tracking microbiome composition over weeks to months after transplant reveals the actual trajectory: whether donor-derived diversity persists and stabilizes into a new, more resilient community state, or whether the recipient's microbiome gradually drifts back toward its pre-transplant, disrupted baseline.
Community composition immediately after FMT is often unstable — early colonizers can be displaced over subsequent weeks as competition (Stage 3) continues to resolve. A single sample taken a few days post-transplant may overstate or understate eventual engraftment. Longitudinal sampling — repeated composition measurements over an extended follow-up window — is what allows the trajectory itself, not just a point-in-time state, to be assessed.
Simplifying considerably, longitudinal monitoring tends to reveal one of two broad patterns: donor-derived diversity increases and then stabilizes into a new, more diverse community state that persists on subsequent follow-up; or the initial gains erode and the community gradually reverts toward something resembling the pre-transplant disrupted baseline, sometimes accompanied by recurrence of the original clinical problem.
Which pattern dominates in a given case depends on the same factors already discussed — pre-transplant disruption severity, strain-level competitive outcomes, and taxon-specific engraftment durability — compounding over time rather than resolving at a single moment.
This simulator's "diversity trajectory" and "durability outlook" metrics are illustrative simplifications of this longitudinal concept, driven by the days-since-FMT and disruption-severity controls — not a predictive clinical tool.
Longitudinal composition tracking is informational — it characterizes how engraftment is trending — and in practice informs decisions such as whether a repeat FMT or additional intervention may be warranted if reversion is detected. Framing monitoring as an ongoing feedback loop, rather than a one-time confirmation, is consistent with the fact that engraftment itself is a gradual, competitive, and only partly predictable process.