{"id":"5e57f7a6-2983-451b-9555-f2b009bbc02b","arxiv_id":"2607.25573","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For trait-structured selection-mutation Lotka-Volterra systems, long-run total population sizes obey the classical generalized Lotka-Volterra ODE, giving rigorous coexistence/extinction and mutation-rate selection results.","lead":"This paper proves that a multi-species model where populations evolve by mutation and selection, with competition depending only on total numbers, has the same long-run outcomes as the classical Lotka-Volterra equation. It yields rigorous coexistence and extinction criteria and shows that slow mutators win in constant environments, while intermediate mutation rates can win in shifting environments.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 2.1(b)/2.2(b) omit the global-stability hypothesis needed for Lemma 3.4; unique positive equilibria need not attract (May–Leonard).","rationale":"The reader's weakest assumption is exactly the load-bearing gap I find. The proof of Theorem 2.1 invokes Lemma 3.4, whose hypothesis 'y* is a global attractor' is not implied by the conditions listed in Theorem 2.1(b). The May–Leonard counterexample is a standard phenomenon in competitive Lotka–Volterra systems and falls squarely within the theorem's scope, so the theorem statement is overbroad. However, the core reduction—exponential decay of α(t), boundedness of ρ(t), and application of the asymptotically autonomous theory—remains coherent once the missing hypothesis is supplied. Cases (a) and (c) are not affected by this gap. Therefore the reader's CONDITIONAL verdict is appropriate; I see no reason to strengthen or weaken it, provided the theorem statements are corrected as recommended.","tokens_in":21109,"tokens_out":6430,"duration_ms":59514,"concrete_test":"Solve the 3-species GLV (5) with λ = (-1,-1,-1) and the cyclic competitive matrix A = [[1, 0.9, 1.2], [1.2, 1, 0.9], [0.9, 1.2, 1]]. Here all a_ij > 0 and the unique positive equilibrium is ρ* = (1/3.1, 1/3.1, 1/3.1). Simulate from a small perturbation of the interior equilibrium and from a perturbation of a boundary equilibrium for large T. Since α = 0.9 < 1, β = 1.2 > 1, and α + β = 2.1 > 2, May–Leonard theory predicts that solutions do not converge to ρ* but approach a heteroclinic cycle. If the numerical trajectory confirms this, Theorem 2.1(b) as stated is false and must be amended to require ρ* to be a global attractor of (5).","verdict_should_be":"UNCHANGED","load_bearing_attack":"The reduction to an asymptotically autonomous GLV is plausible, but the proof of Theorem 2.1 relies on Lemma 3.4, which requires the equilibrium ρ* of the autonomous system (5) to be a global attractor. In case (b), the theorem assumes only that ρ* is the unique positive equilibrium and λ_i < 0, with all interactions competitive (a_ij > 0) or all mutualistic. For N ≥ 3 competitive Lotka–Volterra systems, uniqueness of the positive equilibrium does not imply global attractivity: standard May–Leonard systems have a unique interior equilibrium and a heteroclinic cycle on the boundary, so solutions do not converge to ρ*. Remark 2.3 acknowledges this gap ('the existence of such a state is an assumption' for case (b)), but the formal theorem statements do not include it. Since the central claim for competitive multispecies coexistence depends on this step, Theorem 2.1(b) and Theorem 2.2(b), as stated, overclaim. The issue is fixable by explicitly adding a global-attractor or Lyapunov-stability hypothesis for ρ* in the b) case.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies a system of N Lotka-Volterra type integro-differential equations with mutation diffusion in a continuous trait, trait-dependent growth rates, but trait-independent pairwise interactions. By transforming to moving-frame principal eigenfunctions, the authors derive an ODE system for total population masses ρ(t) that is a perturbation of a generalized Lotka-Volterra system by an exponentially decaying term α_i(t). The main results (Theorems 2.1 and 2.2) claim that under (a) Lyapunov diagonal stability, (b) purely competitive or mutualistic interactions with a unique coexistence equilibrium, or (c) uniform competition, the masses converge to the equilibrium of the limiting ODE (or to the single survivor in case (c)), and the normalized trait distributions converge to the associated principal eigenfunctions. The paper also discusses mutation-rate selection in static and shifting environments and includes numerical simulations.","tokens_in":21337,"tokens_out":12833,"duration_ms":116256,"significance":"If the main theorems hold, the paper provides a rigorous and useful reduction: the long-time dynamics of a complex trait-structured PDE are determined by the equilibria of a well-studied GLV ODE. The proof strategy is transparent, relying on principal eigenvalues, exponential convergence for linear parabolic equations, and Strauss-Yorke theory for asymptotically autonomous systems. The concrete claims about minimal versus intermediate mutation rates are falsifiable and biologically relevant. However, as detailed below, the purely competitive case is not justified by the stated assumptions, so the scope of the main theorems must be narrowed before the paper can be accepted.","major_comments":[{"comment":"The proof of case (b) invokes Lemma 3.4, which requires the equilibrium ρ* of the limiting GLV system (5) to be a global attractor. The theorems only assume that ρ* is the unique positive equilibrium and that interactions are all competitive (a_ij>0) or all mutualistic (a_ij≤0 for i≠j). For N≥3 competitive Lotka-Volterra systems, uniqueness of a positive equilibrium does not imply global attractivity; standard May-Leonard systems possess a unique interior equilibrium together with a heteroclinic cycle, so solutions do not converge to ρ*. Remark 2.3 already concedes that 'the existence of such a state is an assumption' for case (b), but this hypothesis is not included in the formal theorem statements. Consequently Theorem 2.1(b) and Theorem 2.2(b) overclaim. The fix is to add a global-asymptotic-stability hypothesis on ρ* for (5) in case (b).","section":"Theorem 2.1(b), Theorem 2.2(b), Lemma 3.4, Remark 2.3"},{"comment":"The derivation of the reduced ODE has a sign error. Differentiating (9) gives dρ_i/dt = ρ_i(-λ_i - ∑_j a_ij ρ_j - α_i(t)) with α_i defined as -(d˜ρ_i/dt)/˜ρ_i. The paper writes +α_i(t) in (28) and in the statement of Lemma 2.1. Since α_i decays exponentially, the asymptotic conclusions are unaffected, but the reduced equation as stated is inconsistent with (9), and the boundedness estimates in Lemma 3.3 need to be adjusted (using |α_i|). Please correct the sign or redefine α_i accordingly.","section":"Eq. (28) and Lemma 2.1"}],"minor_comments":[{"comment":"The summation index in the interaction term is printed as i, the same as the species index; it should be j.","section":"Eq. (2) and (3)"},{"comment":"The statement defines ρ_i(x) = ∫_{Ω_i} u_i(x,t) dx; the left side should be ρ_i(t).","section":"Theorem 2.1"},{"comment":"The display '∥˜u_i(·,t)−Kp_i∥_{L∞(Ω_i)}, |d˜ρ_i/dt| < e^{A−Bt}' is ambiguous; both quantities should each be bounded by e^{A−Bt}.","section":"Lemma 3.2"},{"comment":"There is a typo: 'w_i p_0' should be 'w_i p_i'; also the boundary assertion should be p_i>0 on the closure, not only on ∂Ω_i, if that is what is used.","section":"Lemma 3.1"},{"comment":"The estimate t ≤ ϵ/diam(C) has inconsistent units (diam(C) is in state-space units). The bound and its derivation should be revisited.","section":"Lemma 4.2"},{"comment":"The argument that the perturbed system has no periodic orbits assumes α(t+T)≠α(t) for every T; if the perturbation is e^{-t}, this is true, but the proof should state this explicitly. Also the claim that the second part is 'straightforward' is too terse.","section":"Section 4, Lemma 4.1"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Worth a serious referee, but two theorem statements need to be repaired.\n\nThe genuinely new result here is the multi-species selection-mutation model with trait-independent interactions, and the clean reduction of the PDE mass dynamics to the GLV ODE. The exponential decay of α_i(t) is convincingly established, and the use of the Strauss–Yorke asymptotic-autonomy theorem is appropriate. The uniform-competition case proves weakest-mutator exclusion for arbitrary N — that does settle an open problem in the spatial-diffusion literature. The mutation-rate selection corollary in a shifting environment is a nice payoff.\n\nNow the soft spot, and it is the same one the stress-test flags. Theorems 2.1(b) and 2.2(b) claim convergence to ρ* for all-competitive or all-mutualistic interactions under only uniqueness of ρ*. The proof, however, applies Lemma 3.4, which requires ρ* to be a global attractor of the limit system. For N≥3 competitive Lotka-Volterra, uniqueness does not imply global attractivity (May–Leonard cycles are the standard counterexample). Remark 2.3 acknowledges this by saying that for case (b) 'the existence of such a state is an assumption,' but the formal statements do not include it. So the theorems as written overclaim. This is easy to fix: state the global-attractor hypothesis explicitly in case (b) (or replace it with a known sufficient condition). The mutualistic case is already covered by the Lyapunov argument, so the issue is restricted to the purely competitive case.\n\nMinor quibbles: Section 4 is somewhat tangential, and Lemma 4.2's bound looks hand-wavy (the cycle diameter appears to be used as a global Lipschitz constant for g, which is not obviously valid). The numerics illustrate the results but are not fully reproducible from the text. Neither affects the main theorems.\n\nVerdict: conditional acceptance in spirit. The reduction is sound, the gap is real but fixable, and the uniform-competition result alone is worth publishing. I would cite this work and would send it to peer review. The authors are clearly thinking carefully; the acknowledgment in Remark 2.3 shows they know the issue even if the theorem statements do not.","headline":"Worth a serious referee, but two theorem statements need to be repaired.","tokens_in":21848,"tokens_out":4162,"would_cite":true,"duration_ms":38952,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35Q92","35K57","92D25","92D15"],"pacs":[],"model":"deepseek-v4-flash","headline":"For trait-structured interacting populations, the long-term outcome equals that of the generalised Lotka-Volterra equation with effective growth rates given by principal eigenvalues.","keywords":["integro-differential Lotka-Volterra","trait-structured populations","selection-mutation","principal eigenvalue","asymptotically autonomous systems","competitive exclusion","mutation rate evolution","coexistence"],"falsifier":"Construct a three-species purely competitive matrix A with a unique positive equilibrium that is not globally attracting (for instance, a heteroclinic cycle), set up system (4), and show numerically or analytically that the total masses do not converge to that equilibrium but instead follow the cycle; this would contradict Theorem 2.1(b).","tokens_in":1425,"feed_emoji":"🧬","tokens_out":6144,"duration_ms":83809,"temperature":0.7,"pith_summary":"The paper studies N populations structured by a continuous trait with mutation and competition depending only on total population sizes. Its central claim is that, under standard interaction assumptions, the eventual survival and trait distributions are exactly those of the classic generalised Lotka-Volterra ODE, with growth rates set by principal eigenvalues. A corollary is that in a constant environment the smallest mutation rate is selected, while in a shifting environment an intermediate mutation rate can be selected. This matters because it reduces a hard integro-differential problem to a well-understood finite-dimensional system.","feed_headline":"Trait-structured ecology reduces to a classic ODE","feed_subtitle":"Total population masses and trait shapes converge to Lotka-Volterra equilibrium and eigenfunctions.","key_machinery":"A variable transformation removes competition and time-dependence from the reaction term, reducing the nonlinear PDE to N decoupled linear parabolic equations whose solutions converge exponentially to principal eigenfunctions p_i. The accompanying principal eigenvalues λ_i become effective growth rates in the limiting ODE. The total-mass equation then appears as a generalised Lotka-Volterra system with an exponentially decaying perturbation, so an asymptotically-autonomous theorem transfers the attractor of the limiting ODE to the full PDE.","core_discovery":"Theorems 2.1 and 2.2 show that when the interaction matrix is Lyapunov diagonally stable, purely competitive, purely mutualistic, or uniform competition, the total population masses converge to the unique positive equilibrium of the associated generalised Lotka-Volterra system (or to the single survivor with the minimal λ_i/a_i), and the normalized trait distributions converge to the principal eigenfunction. The proof shows the mass equation is the GLV dynamics plus an exponentially decaying perturbation, then applies a standard asymptotically-autonomous result.","pith_inferences":["A broader principle suggested by the paper is that any trait-structured system with trait-independent interactions inherits the attractor structure of its mean-field ODE, which may extend to periodic environments where the limiting system is periodic.","The numerical examples show early-time discrepancies between the PDE mass and the GLV solution, implying that a time-delay correction could keep trajectories close for finite times, a testable prediction for empirical population data.","The ratio λ_i/a_i acts as a fitness landscape; experiments measuring mutation rates under controlled environmental shifts could test the predicted transition from minimal to intermediate mutation rates.","Because the perturbation decays exponentially, the approach yields explicit rates of convergence to the GLV equilibrium, quantifying how quickly trait structure becomes irrelevant."],"forward_implications":["Long-term coexistence or extinction can be predicted from the simpler generalised Lotka-Volterra ODE, without simulating the full PDE to long times.","In a constant environment, the population with the smallest mutation rate competitively excludes all others, for any number of species.","With shifting optimal traits, the principal eigenvalue has an interior minimum, so an intermediate mutation rate is selected when the environment shifts fast enough.","The eventual trait distribution is the principal eigenfunction of the mutation-selection operator, giving an explicit phenotypic steady state.","The reduction extends to anisotropic trait-dependent diffusion and to shifting-optimum environments, making the GLV-equivalence robust."],"fun_headline_variants":["Trait-independent interactions yield classic Lotka-Volterra long-term","Environment flips optimal mutation rate","Total population masses follow GLV, traits follow eigenfunction","Selection-mutation model reduces to ODE at long times","Minimal mutation static, intermediate dynamic"],"cache_read_input_tokens":23168,"weakest_assumption_plain":"The conclusions for purely competitive or mutualistic interactions require the limiting generalised Lotka-Volterra equation to have a globally attracting coexistence equilibrium, a property the paper assumes without proof and which can fail for N≥3 even when a unique positive equilibrium exists.","fun_headline_variants_meta":{"raw":{"variants":["Trait-independent interactions yield classic Lotka-Volterra long-term","Environment flips optimal mutation rate","Total population masses follow GLV, traits follow eigenfunction","Selection-mutation model reduces to ODE at long times","Minimal mutation static, intermediate dynamic"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00029,"raw_usage":{"total_tokens":1485,"prompt_tokens":645,"completion_tokens":840,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":389,"completion_tokens_details":{"reasoning_tokens":766}},"tokens_in":389,"tokens_out":840,"duration_ms":9540,"temperature":1.0,"reasoning_tokens":766,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T02:02:17.143873+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Construct a three-species purely competitive matrix A with a unique positive equilibrium that is not globally attracting (for instance, a heteroclinic cycle), set up system (4), and show numerically or analytically that the total masses do not converge to that equilibrium but instead follow the cycle; this would contradict Theorem 2.1(b).","supporting_citations":[],"review_version":1}