{"id":"6b1e2ea2-83ad-4a4c-a516-60f2aff7f454","arxiv_id":"1908.02316","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":8,"one_line_summary":"A five-equation chemostat model of cross-feeding microbes shows stable coexistence over broad trait differences and a continuous mutualism-to-parasitism shift when external nutrients are supplied.","lead":"Two microbial species that feed on each other's waste products can survive together in a well-mixed growth vessel, and a new mathematical model shows the relationship is surprisingly tolerant of differences in growth rates and other traits. It also shows that adding one of the exchanged nutrients from outside can gradually turn the relationship from mutualistic to parasitic in the same pair of species.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Reported steady states do not satisfy the model's own equations: at Table 3's r2=1 row, β≈0.053 h^-1, not ω=0.1 h^-1, so the numerical support for both central claims collapses.","rationale":"The reader's CONDITIONAL verdict focused on the unmeasured efflux rate and lack of experimental validation. My check identifies a more fundamental issue: the numerical states used to support the headline claims violate the model's own steady-state equations for the stated parameters. The reader's remark that the Section 4 algebra checks out is formally true, but the tabulated solutions do not satisfy that algebra at the reported parameter values. I am not claiming the model is incapable of producing the qualitative phenomena after correction; however, as submitted, the paper provides no valid numerical demonstration of either trait-robust stability or the mutualism-to-parasitism spectrum. Because no code or data are provided, the failure cannot be checked or repaired by the reader. This warrants moving the verdict from CONDITIONAL to REJECT for the present manuscript, or at minimum UNVERDICTED pending corrected tables and a reproducibility artifact.","tokens_in":14637,"tokens_out":24321,"duration_ms":237806,"concrete_test":"Independently recompute the symmetric steady state from Eqs. (2.1)–(2.4) with Table 1 parameters and r1=r2=1. First, substitute the Table 3 row r2=1 values into Eq. (2.2): if β≈0.053 h^-1 ≠ ω, that row is not a steady state. Then integrate the ODEs numerically from N1=N2=1e8 cells/ml, R=2e12 fg/ml, and M1=M2=0; the converged coexistence equilibrium should have R≈1.1e10 fg/ml, M≈2.9e12 fg/ml, and N≈9.9e8 cells/ml, not the tabulated R=5.6e9. If the solver reaches a different equilibrium, the tables are invalid and all parameter scans in Sections 5 and 6 must be rerun.","verdict_should_be":"REJECT","load_bearing_attack":"The most load-bearing problem is more specific than the unmeasured efflux rate: the tabulated equilibria are not equilibria of the stated model. With Table 1 parameters, K=10 µg/ml = 1e10 fg/ml, L=100 µg/ml = 1e11 fg/ml, Rin=2 g/l = 2e12 fg/ml, r1=r2=1 h^-1, and ω=0.1 h^-1, the Table 3 row for r2=1 lists N1=N2≈1e9, R∞=5.6e9 fg/ml, and M1∞=M2∞=2.9e12 fg/ml. Substituting into Eq. (2.1), β1 = β2 = (2.9e12/(2.9e12+1e10)) × (5.6e9/(5.6e9+1e11)) × 1 ≈ 0.053 h^-1, far below ω=0.1 h^-1. Hence dN/dt=(β−ω)N≈−0.047N and the reported state is not stationary. Equivalently, the paper's own Eq. (4.4) requires r1/ω × R∞/(R∞+L1) > 1 for a positive M2∞; the table gives 10 × 0.053 = 0.53 < 1, i.e. a negative metabolite concentration. The tabulated R and M values match the solution for r≈2 h^-1 rather than the stated r=1 h^-1. Tables 2–8, which constitute the entire evidence for trait-robust stability and the symbiotic spectrum, are therefore not derived from the model as specified. This is an internal consistency failure, not a matter of parameter calibration or absent experimental data.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper derives a five-ODE chemostat model of disjunctive metabolic symbiosis, in which each of two microbial species grows on a shared limiting resource and on a metabolite excreted by the other species. The author writes down Monod-type growth functions, defines mass-balance equations for populations, resource, and two exchanged metabolites, and then studies non-trivial steady states and their stability. The main claims are that the coexistence equilibrium is robust to large differences in intrinsic growth rates and metabolite-related traits (Section 5), and that varying the external influx concentration of one metabolite shifts the interaction continuously from mutualism through commensalism to parasitism (Section 6). These claims are supported by tables of numerically computed equilibria and Jacobian spectra, and by phase-plane plots of initial-condition dependence.","tokens_in":15052,"tokens_out":3820,"duration_ms":38665,"significance":"If the numerical results were correct, the paper would provide a mechanistic, parameter-based explanation of two biologically interesting phenomena: trait-robust coexistence in syntrophic pairs and environment-dependent shifts along a mutualism–parasitism continuum. The model construction itself is transparent and largely parameter-measurable, and the Jacobian in Eq. (4.6) is correctly dimensioned. However, the paper's evidence is entirely numerical, and the numerical equilibria reported in Tables 2–8 do not satisfy the model's own steady-state equations. Because both central claims rest on these tables, the significance of the manuscript as it stands is not established.","major_comments":[{"comment":"The reported non-trivial steady states are not equilibria of the stated model. Using Table 1 parameters (K1=K2=10 µg/ml = 1e10 fg/ml, L1=L2=100 µg/ml = 1e11 fg/ml, r1=r2=1 h^-1, ω=0.1 h^-1), the Table 3 row for r2=1 lists N1=N2≈1.0e9 cells/ml, R(∞)=5.6e9 fg/ml, and M1(∞)=M2(∞)=2.9e12 fg/ml. Substituting into Eq. (2.1) gives β1=β2≈(2.9e12/(2.9e12+1e10))·(5.6e9/(5.6e9+1e11))·1 h^-1≈0.053 h^-1, which is less than ω=0.1 h^-1. Hence dN_i/dt=(β_i−ω)N_i<0 at the reported state, so it is not stationary. The same inconsistency appears in Table 2: with R(∞)=5.6e9 and r1=1, β1 cannot reach 0.1 h^-1 regardless of M2. The tabulated R(∞) values correspond to a resource Monod factor of about 0.053, which would require an intrinsic growth rate near 1.9 h^-1 to balance washout, not r=1 h^-1.","section":"Section 4, Tables 2 and 3"},{"comment":"The analytical steady-state formula is itself inconsistent with Table 3. Equation (4.4) requires (r1/ω) · R(∞)/(R(∞)+L1) > 1 for a positive M2(∞). For the r2=1 row, R(∞)=5.6e9, L1=1e11, r1/ω=10, so the product is 10 × 0.053 ≈ 0.53 < 1, which yields a negative metabolite concentration, not the positive M2(∞)=2.9e12 in the table. This is not a rounding issue; the reported state violates the model's own necessary condition for a feasible equilibrium.","section":"Section 4, Eq. (4.4)"},{"comment":"The symbiotic spectrum is computed as the ratio of coculture equilibria to monoculture equilibria, but both sets of equilibria are obtained from the same non-stationary states. Since the Table 3 equilibria are not equilibria, the ratios in Tables 7 and 8 do not measure the effect of coculture on equilibrium population densities. The claim that increasing M2,in shifts the interaction from mutualism through commensalism to parasitism is therefore unsupported by the numerical evidence presented.","section":"Section 6, Tables 7 and 8"},{"comment":"The constant metabolite efflux rate is a load-bearing parameter that is openly estimated: 'Due to the lack of research literature on the topic, the efflux rate had to be estimated to be 300 fg/h.' All equilibrium values, stability ranges, and interaction-type transitions in Tables 2–8 depend on this value, and no sensitivity analysis with respect to ϵ1, ϵ2 is provided. This would be a serious caveat even if the equilibria were internally consistent; here it compounds the correctness problem.","section":"Section 3"}],"minor_comments":[{"comment":"The word 'infering' should be 'inferring' in the description of Mee et al. (2014).","section":"Introduction"},{"comment":"The sentence 'This can be explained by the fact that increasing the external metabolite concentrations will increase the realised growth rates of the two populations and, therefore, allow them to grow faster than the washout rate ω' repeats 'state' as 'state state' in the preceding sentence: 'converge to the stable state state of non-extinction.'","section":"Section 4"},{"comment":"The Jacobian of the extinction state with added metabolites is described as 'a lower diagonal matrix'; it is lower triangular, not lower diagonal in the usual sense.","section":"Section 4, Eq. (4.7)"},{"comment":"The captions state that 'All values are rounded to the first decimal place,' but the entries are given with one or two significant figures and orders of magnitude; this phrasing is inaccurate and should be corrected.","section":"Tables 3–6 and 9–10"},{"comment":"The table for K2 lists values of M2(∞) that decrease with K2 until the K2=1e10 row, after which M1(∞) and M2(∞) swap roles; the caption does not note this symmetry, which may confuse readers.","section":"Section 5.2, Table 6"},{"comment":"Some references are cited by author-year in the text (e.g., Miller, 1972; Raetz, 1996) but the reference list entries are not all formatted consistently; for example, Raetz (1996) is listed without a full citation title.","section":"References"}],"recommendation":"reject","confidential_remarks":"The central numerical support for both headline claims is internally inconsistent: the tabulated equilibria do not satisfy the model's own equations. Because the manuscript contains no experimental data and no analytical stability proof that would salvage the qualitative conclusions, the errors cannot be fixed by local edits; the numerical sections would need to be recomputed and revalidated. I recommend rejection, though the model formulation itself could serve as a starting point for a future substantially revised manuscript."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's the thing: the paper presents a five-ODE chemostat model of disjunctive syntrophy and, on the symbolic side, much of Section 4 is sound. The extension to include a shared limiting resource and asymmetric traits is legitimate, and the Jacobian is correctly built. The qualitative idea—that the interaction can move along a mutualism-to-parasitism spectrum—fits with earlier experimental work, so the motivation is fine.\n\nBut the numerical results do not hold up. Take Table 3's r2=1 row: N1=N2≈1e9, R∞=5.6e9 fg/ml, M1=M2=2.9e12 fg/ml. With Table 1 parameters, the effective growth rate β from Eq. (2.1) is about 0.053 h−1, well below ω=0.1 h−1. So dN/dt=(β−ω)N is −0.047N, not zero. Similarly, Eq. (4.3) gives R∞≈0 if N1=N2=1e9 and α=1000 fg/cell, not the reported 5.6e9. The same problems appear in the other tables, including the 'stable coexistence' row in Table 2. These are not rounding issues; they are states that are not equilibrium states of the stated model. The tables are the only evidence for both central claims—robustness to trait asymmetry and the resource-mediated spectrum—so those claims are unsupported.\n\nThe unmeasured efflux rate (ε=300 fg/h) is a soft spot, but it is secondary. Even with a perfect parameter set, the reported numbers would still violate the model. The stability analysis also relies on sampled eigenvalues rather than derived conditions, which is weaker.\n\nIf the numerics were redone, the qualitative conclusions might still emerge. But as written, the quantitative support collapses. The symbolic equilibrium relations could be of use to someone building on this model, but I would not cite the stability tables.\n\nMy recommendation: desk reject or ask for major revision with reanalysis. A serious referee would likely catch this and reject anyway; better to have the author rerun the calculations first. If it comes back with corrected tables and a measured or properly bounded efflux rate, then it would be worth a real review. As it stands, no.","headline":"The model formulation is sensible, but the numerical tables that carry the main claims are internally inconsistent with the paper's own equations.","tokens_in":15594,"tokens_out":6658,"would_cite":false,"duration_ms":61929,"reading_group":"no","serious_thinker":"no","would_accept_peer_review":false},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper claims that a mechanistic five-equation chemostat model of reciprocal microbial cross-feeding produces a stable coexistence equilibrium across wide trait asymmetries, and that the species interaction shifts continuously from…","keywords":["disjunctive symbiosis","syntrophy","cross-feeding","chemostat","mechanistic mathematical model","Monod kinetics","symbiotic spectrum","metabolic interdependence"],"falsifier":"Measure the real efflux rates $\\epsilon_1,\\epsilon_2$ (and the yield factors and Monod constants) for a specific auxotrophic pair, run the chemostat model with those numbers, and compare the predicted coexistence densities and the $M_{2,\\mathrm{in}}$ value at which the coculture-to-monoculture ratio crosses unity; if the measured efflux rate is outside the stable range, or if the interaction spectrum does not shift monotonically with $M_{2,\\mathrm{in}}$, the paper's quantitative claims fail. A crisper test is the bistability prediction: inoculate below the critical initial density shown in the phase-space diagrams and check whether the culture reproducibly collapses to extinction, or inoculate above it and check that coexistence is reliably attained.","tokens_in":14402,"feed_emoji":"🦠","tokens_out":9877,"duration_ms":89367,"temperature":0.7,"pith_summary":"The paper proposes a mechanistic mathematical model of disjunctive microbial syntrophy--two species that each live on a metabolite released by the other--as five coupled differential equations for the chemostat: two population densities, one shared limiting resource, and the two exchanged metabolites. All parameters are intended to be directly measurable biological quantities, such as growth rates, Monod constants, and cellular efflux and yield factors, rather than fitted phenomenological coefficients. Using this model, the author argues that the symbiotic equilibrium is stable across wide trait differences: it persists when the symbionts differ strongly in intrinsic growth rate, efflux rate, metabolite requirement, or substrate affinity. The model's central result is that the interaction type is environmentally tunable--increasing the external concentration of one exchanged metabolite pushes the same species pair continuously from obligate mutualism through commensalism to parasitism, with the availability of the shared resource modulating where that transition occurs.","feed_headline":"External metabolite supply turns mutualists into parasites","feed_subtitle":"Five-equation chemostat model predicts interaction type is set by medium, not just species traits.","key_machinery":"The central mechanism is the five-dimensional system (Eqs. 2.2-2.4) for population densities $N_1,N_2$, the shared resource $R$, and the two exchanged metabolites $M_1,M_2$. Effective growth rates are products of two Monod terms, one for the partner-supplied metabolite and one for the limited resource. The steady-state analysis uses $\\beta_1=\\beta_2=\\omega$ at equilibrium, yielding algebraic equations (4.3-4.5) that link equilibrium metabolite levels and population densities; stability is decided by the eigenvalues of the $5\\times5$ Jacobian. The interaction spectrum is quantified by dividing each species' equilibrium density in coculture by its density in monoculture, so ratios above, near, or below 1 mark mutualism, commensalism, and parasitism, respectively.","core_discovery":"On its own terms, the paper's central claim is that a mechanistic chemostat model of reciprocal cross-feeding reproduces two observed ecological phenomena. First, the nontrivial equilibrium of two metabolically interdependent microbes remains stable over a wide range of trait asymmetries: intrinsic growth rates $r_2$ between roughly $0.11$ and $100\\,\\text{h}^{-1}$, metabolite efflux rates down to $1\\,\\text{fg}/(\\text{cell}\\cdot\\text{h})$, and Monod constants spanning many orders of magnitude. The author interprets this as the system compensating for differences in biological traits. Second, the interaction type is not fixed: raising the influx concentration $M_{2,\\mathrm{in}}$ of the metabolite that species 1 requires makes the coculture-to-monoculture ratio for species 1 fall from effectively infinite (obligate mutualism) through roughly unity (commensalism) to below unity (parasitism), while species 2 always benefits from the partnership. The paper also shows that raising the influx $R_{\\mathrm{in}}$ of the shared resource moves the mutualism-to-parasitism transition to higher $M_{2,\\mathrm{in}}$, and that the extinction state is always stable, so reaching the symbiotic equilibrium requires sufficiently large inocula.","pith_inferences":["If the model's spectrum result is general, then reported interaction types in cross-feeding experiments should be published with the exact medium composition, because the same pair would be classified differently at different external metabolite concentrations.","The bistability between extinction and coexistence suggests a practical route to domesticating 'unculturable' symbionts: larger inocula or metabolite supplementation might push the culture into the coexistence basin, a concrete extension of the paper's own remark about unculturable taxa.","The paper's closing argument on evolutionary erosion implies that long-term stability requires some fitness benefit for metabolite release; coupling these ODEs to adaptive dynamics of the efflux rates $\\epsilon_1,\\epsilon_2$ would be a natural next step.","The compensation of trait asymmetries predicts that in synthetic cross-feeding pairs, a fast-growing symbiont will dominate numerically but not destroy the partnership, which could be tested by measuring abundance ratios across strains with different $r$ values."],"forward_implications":["A stable syntrophic consortium can be assembled from strains whose intrinsic growth rates differ by up to two orders of magnitude, so close trait matching is not a precondition for coexistence.","At external metabolite influx around $10\\,K_1$, the interaction becomes commensal, and above that the first symbiont is harmed by its partner's presence; the relationship is therefore a product of the environment, not a fixed property of the pair.","Raising the supply of the shared limiting resource delays the mutualism-to-parasitism transition, implying that richer media can keep a partnership mutualistic at metabolite concentrations that would otherwise make it parasitic.","Because extinction is always a stable steady state, starter culture sizes determine whether a syntrophic consortium establishes; low inocula collapse, while sufficiently large inocula reach coexistence.","Since all parameters are experimentally accessible, the model yields quantitative predictions that can be tested directly in chemostat co-cultures of synthetic auxotrophic strains."],"supporting_citations":[{"why":"Supplies the growth-kinetics form linking growth to metabolite and resource concentrations.","marker":"Monod, 1949"},{"why":"Provides the four-equation forced-symbiosis model that the present five-equation model generalizes to non-identical symbionts.","marker":"Kerner et al., 2012"},{"why":"Gives the experimental auxotrophic E. coli cross-feeding system that motivates and grounds the model's biology.","marker":"Mee et al., 2014"},{"why":"Demonstrates the experimental yeast-Lactobacillus 'symbiotic spectrum' that the model aims to reproduce mechanistically.","marker":"Megee et al., 1972"},{"why":"Shows experimentally that resource availability reshapes the interaction spectrum, which the model's continuous-spectrum claim echoes.","marker":"Hoek et al., 2016"}],"fun_headline_variants":["Metabolite dose flips microbe symbiosis from mutualism to parasitism","Stable microbe partnership despite trait differences, model shows","External chemicals decide if microbes are friends or foes","Chemostat model: medium tunes microbe interaction spectrum"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The entire quantitative output rests on the assumption that both symbionts export their exchanged metabolite at a constant rate of 300 fg per cell per hour, a value the author states had to be estimated because no research literature exists on the topic.","fun_headline_variants_meta":{"raw":{"variants":["Metabolite dose flips microbe symbiosis from mutualism to parasitism","Stable microbe partnership despite trait differences, model shows","External chemicals decide if microbes are friends or foes","Chemostat model: medium tunes microbe interaction spectrum"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000286,"raw_usage":{"total_tokens":1658,"prompt_tokens":900,"completion_tokens":758,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":516,"completion_tokens_details":{"reasoning_tokens":689}},"tokens_in":516,"tokens_out":758,"duration_ms":7809,"temperature":1.0,"reasoning_tokens":689,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:49:14.036369+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the real efflux rates $\\epsilon_1,\\epsilon_2$ (and the yield factors and Monod constants) for a specific auxotrophic pair, run the chemostat model with those numbers, and compare the predicted coexistence densities and the $M_{2,\\mathrm{in}}$ value at which the coculture-to-monoculture ratio crosses unity; if the measured efflux rate is outside the stable range, or if the interaction spectrum does not shift monotonically with $M_{2,\\mathrm{in}}$, the paper's quantitative claims fail. A crisper test is the bistability prediction: inoculate below the critical initial density shown in the phase-space diagrams and check whether the culture reproducibly collapses to extinction, or inoculate above it and check that coexistence is reliably attained.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the four-equation forced-symbiosis model that the present five-equation model generalizes to non-identical symbionts."},{"cited_title":"T., Collins, J","cited_arxiv_id":null,"evidence_quote":"Gives the experimental auxotrophic E. coli cross-feeding system that motivates and grounds the model's biology."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Demonstrates the experimental yeast-Lactobacillus 'symbiotic spectrum' that the model aims to reproduce mechanistically."},{"cited_title":"A., Axelrod, K., Biancalani, T., Yurtsev, E","cited_arxiv_id":null,"evidence_quote":"Shows experimentally that resource availability reshapes the interaction spectrum, which the model's continuous-spectrum claim echoes."}],"review_version":1}