{"id":"019e56de-e553-4185-be2e-b8af5bf9339b","arxiv_id":"2411.10167","paper_version":1,"verdict":"REJECT","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"high","formal_verification":"none","parameter_count":0,"one_line_summary":"Assuming a gamma distribution for population size at equilibrium, the paper derives quadratic relations between intrinsic growth rate and noise variance, yielding two growth-rate branches it calls alternative stable states.","lead":"This paper derives explicit formulas linking the intrinsic growth rate r of a discrete stochastic logistic population model to the shape parameter k of a gamma distribution and the variance of random perturbations. The authors claim the growth rate has two possible branches, which they interpret as alternative stable population states.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The equilibrium claim rests on moment matching alone; a gamma law is not shown to be invariant, and the map's negative branch makes exact gamma stationarity impossible.","rationale":"Good-faith reading: the paper's algebraic derivation of the moment conditions is internally mostly consistent, and the simplification leading to Eq. (31) checks out. The problem is interpretive: the title and abstract promise stationary distributions, but the 'equilibrium' used is only first- and second-moment equality. For a nonlinear Markov map, moment matching cannot identify a stationary law: the transition operator is not determined by two moments, and alternative distributions with the same moments would pass the test. The support argument makes the failure even starker: because θ<1, a gamma population has positive probability of exceeding 1, and at those states the map is negative, so the next population is negative; a gamma variable is never negative. Hence the claimed 'gamma equilibrium' cannot be an exact stationary distribution. The variance-bound error at Eq. (32) is a separate internal defect, but it is not the main reason for rejecting the central claim. The verdict REJECT stands.","tokens_in":8126,"tokens_out":11666,"duration_ms":113731,"concrete_test":"Set v=0.1, k=1, let ε be lognormal with Eε=1 and Varε=v, and take r=r− from Eq. (39). With θ=(r−1)/(r(1+k)), compute S = r^3(1+v)^3 [1−3θ(k+3)+3θ^2(k+3)(k+4)−θ^3(k+3)(k+4)(k+5)], which equals E[X_{t+1}^3]/E[X_t^3] if X_t~Gamma(k,θ). A stationary gamma law requires S=1; for these parameters S≠1, so the first three moments are not invariant. Repeating for r+ gives a negative S, reinforcing the support failure. This directly falsifies gamma invariance under (2).","verdict_should_be":"REJECT","load_bearing_attack":"Section 2.1 and Eq. (11) define equilibrium by equality of the first two moments of X_{t+1} and X_t. For the nonlinear map (2), equal first two moments are necessary, not sufficient, for a stationary distribution: many distributions share those moments while the full law changes. The paper never proves that the gamma law is invariant, and it cannot: with θ=(r−1)/(r(1+k))<1, a gamma X_t has P(X_t>1)>0, and on that event rX_t(1−X_t)<0, so X_{t+1}<0. A gamma random variable is nonnegative, so X_{t+1} has the wrong support; the exact gamma distribution is therefore not invariant under (2). The two roots in Eq. (39) are moment-calibration curves, not 'alternative stable states.' The paper's own Conclusion concedes that a rigorous framework for stationary distributions is future work. Independently, the feasibility bound is mis-derived: from D=4(k+1)^2[1−v(k+2)/(v+1)]≥0 one gets k≤1/v−1, not k≤1/v−2, so the claimed cap Var(ε)≤0.5 is not supported (the actual condition is v<1).","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the stochastic logistic recurrence X_{t+1}=rX_t(1-X_t)\\epsilon_t with E[\\epsilon_t]=1 and nonnegative perturbation. It assumes that, at equilibrium, X_t follows a gamma distribution with shape k and scale \\theta (Eq. 4), and imposes equality of the first two moments of X_{t+1} and X_t (Eq. 11). From the mean condition it obtains \\theta=(r-1)/(r(1+k)) (Eq. 17), and from the variance condition it derives a quadratic equation for r (Eq. 31). The roots are given in Eq. (39), and the discriminant is used to claim the parameter bounds 0<k\\le 1/Var[\\epsilon_t]-2 and 0\\le Var[\\epsilon_t]\\le 0.5. The paper interprets the two roots r_+ and r_- as alternative stable states and argues that the feasible range of r is the same as for the deterministic logistic map, namely 1<r<3.","tokens_in":8351,"tokens_out":7594,"duration_ms":73858,"significance":"If the gamma law were shown to be an exact stationary distribution of (2), the explicit r-k-variance relation would be a compact and potentially useful result for stochastic logistic population models, and the analogy with the deterministic interval 1<r<3 would be biologically appealing. The algebraic derivation from Eq. (18) to Eq. (31) is internally consistent, and the mean condition leading to Eq. (17) is straightforward. However, the central claim is not established: the paper verifies only two moment equalities, not invariance of the full gamma distribution, and the gamma's positive support is incompatible with the fact that the map (2) produces negative values whenever X_t>1. No code or numerical simulation is provided; the figures plot Eq. (39) and therefore cannot validate the equilibrium claim independently. The paper is better read as a moment-calibration exercise than as a proof of stationary gamma equilibria.","major_comments":[{"comment":"The paper defines equilibrium by equality of the first two moments of X_{t+1} and X_t. For the nonlinear map (2), these equalities are necessary but not sufficient for stationarity of the full distribution. No argument shows that X_{t+1} has the same gamma distribution as X_t, and the paper itself notes immediately after Eq. (10) that X_{t+1} may not be gamma. Consequently, the roots in Eq. (39) are solutions of a moment-calibration equation, not a proof of the existence of a stationary gamma distribution.","section":"Section 2.1, Eq. (11)"},{"comment":"The support of the gamma distribution is incompatible with the map (2). With \\theta=(r-1)/(r(1+k)) and r>1, one has \\theta<1, so a gamma random variable X_t has positive probability of exceeding 1. On the event X_t>1, the quantity rX_t(1-X_t) is negative, and since \\epsilon_t is assumed nonnegative, X_{t+1}<0 with positive probability. A gamma random variable is nonnegative, so X_{t+1} cannot be exactly gamma distributed. This is not a mild approximation issue; the supports are disjoint. The exact stationary gamma claim therefore fails for every r>1.","section":"Section 2.2, Eq. (17)"},{"comment":"The feasibility bound is algebraically incorrect. From D=4(k+1)^2[1 - v(k+2)/(v+1)] with v=Var[\\epsilon_t], the condition D\\ge 0 is equivalent to v(k+1)\\le 1, i.e. k\\le 1/v - 1, not k\\le 1/v - 2. Hence the claimed upper bound Var[\\epsilon_t]\\le 0.5 does not follow; the discriminant allows values such as v=0.6 with k\\le 2/3. Since this bound is used in Section 2.4.2 to restrict the admissible range of k and in Figure 2, the reported parameter range must be corrected.","section":"Section 2.4, Eqs. (32)-(33)"},{"comment":"The interpretation of r_+ and r_- as \"alternative stable states\" is unsupported. The two roots are merely the two solutions of a quadratic moment equation; the paper provides no stability analysis, no argument that trajectories converge to either branch, and no demonstration that the two branches correspond to distinct attracting equilibria. Moreover, r is not an externally specified growth parameter in this construction; it is a function of k and v. The statement that the feasible range matches the deterministic condition 1<r<3 is therefore not a dynamical equivalence but a property of Eq. (39).","section":"Section 2.4, Eq. (39)"},{"comment":"The Conclusion explicitly states that \"a more rigorous mathematical framework is needed to prove the relationships between the parameters of stochastic models and their stationary distributions\" and suggests that beta distributions might be more suitable because they respect the carrying capacity. These statements are in tension with the Abstract's claims that the paper identifies the feasible range of r at equilibrium and establishes alternative stable states. The paper's own caveats thus acknowledge that the central stationarity result is not proven in the current manuscript.","section":"Section 3, Conclusion and Discussion"}],"minor_comments":[{"comment":"The phrase \"analysis the impact\" should be \"analyze the impact,\" and the repeated phrase \"sufficient small\" in Sections 2.4.1 and 2.4.2 should be \"sufficiently small.\"","section":"Abstract and Section 1"},{"comment":"Equation (9) is followed by an empty numbered equation (10); the numbering should be cleaned up so that the display is not followed by a blank line.","section":"Section 2.1, Eqs. (9)-(10)"},{"comment":"Figure 2 is described as containing four panels, but the panels are not labeled and the axis labels and parameter values used in the plots are not specified, which makes the figure difficult to interpret.","section":"Figure 2"},{"comment":"Reference [9] is a Wikipedia article; a standard textbook or peer-reviewed reference for the gamma distribution would be more appropriate in a journal article.","section":"References"},{"comment":"The notation V ar[\\epsilon_t] appears with a space and inconsistent formatting; the right-hand side of Eq. (26) would be clearer with explicit parentheses around the product of the variance fraction, (k+1)^2, and 1/(r-1).","section":"Throughout"},{"comment":"The reference list contains apparent typos: \"Dirersity\" should be \"Diversity\" and \"Stochustic Abundunce\" should be \"Stochastic Abundance.\"","section":"References [19]-[20]"}],"recommendation":"reject","confidential_remarks":"The manuscript is an early draft and the central claim is not supported by the presented mathematics. The moment algebra is likely correct, but the model's support issue and the absence of any stationarity proof are load-bearing, not cosmetic. A resubmission would require either a different stochastic model with nonnegative range, or a clearly labeled moment-closure approximation with independent numerical validation. I see no indication of misconduct; the weakness is scientific."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear [Colleague],\n\nQuick take: this paper is a two-moment closure dressed up as an equilibrium analysis. The author derives explicit formulas for r in terms of the gamma shape k and the noise variance v = Var(ε) — equations (39) and the two-branch structure. The algebra up to the quadratic (31) is straightforward and I didn't find a mistake there. The observation that r is independent of the scale parameter θ is also correct and worth a sentence.\n\nThe problem is the central claim. Equation (11) defines equilibrium by equality of the first two moments of X_{t+1} and X_t. For a nonlinear map like (2), that is necessary but not sufficient for X_{t+1} to have the same distribution as X_t. More concretely, the paper assumes X_t ~ Gamma(k, θ). With θ = (r−1)/(r(1+k)) > 0, the gamma puts positive mass on x > 1. On that event, r x(1−x) < 0, so X_{t+1} is negative. A gamma random variable cannot be negative, so the exact gamma law is never invariant under (2). The paper's own conclusion says \"a more rigorous mathematical framework is needed\" — that is the right caveat, but the abstract and Section 2.4 treat the moment-calibration as a stationary distribution and talk about \"alternative stable states,\" which overstates what has been shown.\n\nThere is also an algebraic slip in the feasibility condition. From D = 4(k+1)^2[1 − v(k+2)/(v+1)] ≥ 0, the condition on k is k ≤ 1/v − 1, not k ≤ 1/v − 2 as in Eq. (32). Consequently the claimed cap v ≤ 0.5 is not supported; the actual necessary condition is v < 1. That is a load-bearing error: the \"maximum variance of the perturbation\" is part of the paper's punchline.\n\nThe literature engagement is fine; the gamma abundance references are appropriate, and the deterministic logistic background is standard. No code or data is needed for a purely analytic result, but with the support problem and the bound error, the result as stated is not a valid stationary-distribution theorem. It could be reframed as a heuristic moment-matching approximation for low-noise regimes, but that is not what the abstract promises.\n\nFor peer review: I would not engage. The stationarity claim fails on its own terms, and the mis-derived bound would have been caught by any careful referee. If the author wants to salvage it, they need to either prove invariance in a suitable approximation sense or explicitly disclaim stationarity and call it what it is: a moment-consistency calculation.\n\nBest,\n\n[You]","headline":"A clean moment-matching calculation that is sold as a stationary-distribution result; the stationarity claim doesn't survive contact with the map's support, and the feasibility bound is off by an algebra step.","tokens_in":8851,"tokens_out":5409,"would_cite":false,"duration_ms":45441,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":false},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["92D25"],"pacs":[],"model":"deepseek-v4-flash","headline":"For a stochastic logistic model with a gamma-distributed equilibrium, the intrinsic growth rate is pinned to two branches, r+ and r−, lying within the same feasible range as the deterministic logistic map.","keywords":["stochastic logistic model","gamma distribution","stationary distribution","moment matching","intrinsic growth rate","alternative stable states","environmental stochasticity","discrete-time population model"],"falsifier":"Simulate the recurrence $X_{t+1}=rX_t(1-X_t)\\epsilon_t$ for a chosen $k$ and $\\mathrm{Var}(\\epsilon)$ with $r$ given by either root in Eq. (39) and $\\theta=(r-1)/(r(1+k))$, using a large ensemble and many time steps; if the empirical distribution of $X_t$ does not converge to $\\mathrm{Gamma}(k,\\theta)$ or the moments drift, the derived equilibrium is not a real stationary distribution.","tokens_in":7875,"feed_emoji":"📈","tokens_out":9033,"duration_ms":72653,"temperature":0.7,"pith_summary":"The paper derives explicit equilibrium conditions for a discrete stochastic logistic model in which the population distribution is assumed to be gamma. It shows that, when the first two moments are stationary, the intrinsic growth rate r must satisfy a quadratic whose two roots, r+ and r−, represent alternative stable states. The result yields a feasible range 1<r<3 that matches the deterministic logistic map, and it links the noise variance and the gamma shape parameter k through the condition that the variance be at most 0.5. The scale parameter θ drops out, so r depends only on the shape of the population distribution and the noise level. A reader would care because this gives a closed-form, parameter-light description of when a fluctuating population can settle into a gamma-shaped equilibrium.","feed_headline":"Gamma equilibrium yields two growth-rate branches","feed_subtitle":"Explicit formulas tie noise variance and distribution shape to r, matching the deterministic r range from 1 to 3.","key_machinery":"The argument's engine is the gamma distribution assumed for $X_t$, combined with the moment-stationarity conditions $E[X_{t+1}]=E[X_t]$ and $\\mathrm{Var}[X_{t+1}]=\\mathrm{Var}[X_t]$ (equation 11). The gamma's moment formula $E[X^n]=\\theta^n\\Gamma(k+n)/\\Gamma(k)$ turns those two conditions into a closed system in $r$, $k$, and $\\theta$. Eliminating $\\theta$ produces a quadratic in $r$, whose discriminant gives the noise-variance ceiling $\\mathrm{Var}(\\epsilon)\\le0.5$ and the shape-parameter bound $k\\le1/\\mathrm{Var}(\\epsilon)-2$. The independence of $\\epsilon_t$ from $X_t$ is what allows the noise to factor out of the expectations, so only the second moment of $\\epsilon_t$ enters.","core_discovery":"The central claim is that for the stochastic logistic recurrence $X_{t+1}=rX_t(1-X_t)\\epsilon_t$ with $E[\\epsilon_t]=1$, an equilibrium population that follows a gamma distribution with shape $k$ and scale $\\theta$ exists only for certain combinations of $r$, $k$, and the noise variance. Assuming the mean and variance are unchanged from $t$ to $t+1$, the paper derives $\\theta=(r-1)/(r(1+k))$ and a quadratic equation for $r$. Its two roots are $r_{\\pm}=(2k+4\\pm(k+1)\\sqrt{1-\\mathrm{Var}(\\epsilon)(k+2)/(\\mathrm{Var}(\\epsilon)+1)})/(k+3)$, with real solutions requiring $0\\le\\mathrm{Var}(\\epsilon)\\le0.5$ and $0<k\\le1/\\mathrm{Var}(\\epsilon)-2$. The two roots $r_+>r_-$ are interpreted as higher- and lower-density alternative stable states, and all feasible $r$ lie in the deterministic logistic range $1<r<3$. Because $\\theta$ cancels from the relation, the growth rate at equilibrium depends on the population's distribution shape and the environmental variance, not on its absolute scale.","pith_inferences":["A natural test of the paper's assumption is to simulate the map with gamma-distributed initial conditions and check whether the distribution at later times remains gamma with the specified parameters; the moment conditions alone do not guarantee invariance of the full distribution.","Because only the second moment of $\\epsilon_t$ enters the formulas, the same equilibrium relations should hold for any noise distribution with mean 1 and the given variance, such as lognormal or a scaled beta, though the actual stationary law would differ.","The two branches $r_+$ and $r_-$ suggest a hysteresis scenario: a population on the high-growth branch might drop to the low-growth branch after a disturbance, and the paper's formulas could be used to test which branch is more resilient to extinction.","Extending the moment-matching approach to a beta distribution, as the paper suggests for models with an explicit carrying capacity, would yield analogous conditions and might reveal how the support boundary changes the two-branch structure."],"forward_implications":["If the equilibrium is gamma-distributed, the intrinsic growth rate $r$ is fixed by the shape parameter $k$ and the noise variance $\\mathrm{Var}(\\epsilon)$; the scale $\\theta$ does not affect $r$.","Environmental noise with $\\mathrm{Var}(\\epsilon)>0.5$ cannot support this kind of gamma equilibrium, and higher noise shrinks the admissible range of $k$.","For fixed $k$ and $\\mathrm{Var}(\\epsilon)$, the two roots $r_+$ and $r_-$ give two distinct equilibria, implying the model can support alternative stable states at different population densities.","The derived feasible range $1<r<3$ matches the deterministic logistic map, so stochastic and deterministic versions share the same stability window.","The formula $\\theta=(r-1)/(r(1+k))$ links the equilibrium mean $k\\theta$ directly to $r$ and $k$, so the population size is fixed once $r$ and $k$ are chosen."],"supporting_citations":[{"why":"Supplies the gamma distribution as a model of population abundance, the key distributional assumption of the paper.","marker":"[4]"},{"why":"Provides the stationary-distribution result for Tribolium populations that motivates the gamma equilibrium assumption.","marker":"[5]"},{"why":"Shows the gamma abundance model applies to steady-state populations, supporting the use of the gamma law.","marker":"[6]"},{"why":"Documents gamma-distributed adult numbers for Tribolium in laboratory populations, the empirical basis for assuming a gamma equilibrium.","marker":"[7]"},{"why":"Establishes the deterministic logistic map dynamics, including the 1<r<3 range of stable nonzero steady states that the paper matches.","marker":"[13]"},{"why":"Standard reference for the stability windows of the deterministic logistic map, used to frame the feasible r range.","marker":"[14]"},{"why":"Provides the classical treatment of the logistic map's dynamics, supporting the comparison with the deterministic case.","marker":"[15]"}],"fun_headline_variants":["Stochastic logistic equilibrium splits growth into two branches","Gamma equilibrium yields pair of stable growth-rate states","Noise variance sets two possible growth rates for gamma equilibrium","Gamma equilibrium reveals alternative stable growth-rate branches"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The entire derivation rests on treating equality of the first two moments from one time step to the next as the definition of equilibrium, without proving that the gamma distribution itself is invariant under the map; many distributions could share those moments without being stationary.","fun_headline_variants_meta":{"raw":{"variants":["Stochastic logistic equilibrium splits growth into two branches","Gamma equilibrium yields pair of stable growth-rate states","Noise variance sets two possible growth rates for gamma equilibrium","Gamma equilibrium reveals alternative stable growth-rate branches"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000904,"raw_usage":{"total_tokens":3895,"prompt_tokens":956,"completion_tokens":2939,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":572,"completion_tokens_details":{"reasoning_tokens":2879}},"tokens_in":572,"tokens_out":2939,"duration_ms":18989,"temperature":1.0,"reasoning_tokens":2879,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T19:53:47.832738+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Simulate the recurrence $X_{t+1}=rX_t(1-X_t)\\epsilon_t$ for a chosen $k$ and $\\mathrm{Var}(\\epsilon)$ with $r$ given by either root in Eq. (39) and $\\theta=(r-1)/(r(1+k))$, using a large ensemble and many time steps; if the empirical distribution of $X_t$ does not converge to $\\mathrm{Gamma}(k,\\theta)$ or the moments drift, the derived equilibrium is not a real stationary distribution.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the classical treatment of the logistic map's dynamics, supporting the comparison with the deterministic case."}],"review_version":1}