{"id":"87cecb59-432a-47cd-ab56-76c234219e04","arxiv_id":"2608.11918","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"A metabolic-scaling forest model predicts a two-regime power-law tree-size distribution, with the crossover set by recruitment and resources, and shows that forest boundaries can cause long-dispersal species to decline across the whole stand.","lead":"This paper builds a mathematical forest where growth, shading, competition, and seed dispersal follow simple metabolic scaling rules. It predicts a two-part power-law size distribution and shows that forest edges can slowly eliminate species that spread seeds over long distances.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The two-exponent stationary solution and the 'exponent as forest condition' readout hinge entirely on the hand-written mortality closures in eqs. (5)-(6); until a mechanistic derivation or sensitivity test shows these closures are faithful, the central claim is not secure.","rationale":"I read the paper in good faith and confirm that the analytical machinery is internally consistent: eq. (14) does follow from the factorized stationary equation (25), and the two power-law regimes are genuine consequences of the two mortality terms. However, the central claim that the observed exponent can be interpreted mechanistically depends on the two mortality closures being faithful representations of real forest competition. The reader's weakest-assumption analysis identified exactly this point, and my reading agrees. The resource-mortality term in eq. (6) is especially consequential because it alone sets the small-size exponent -a and the crossover position; the Methods section does not derive this form from individual-level resource uptake but instead assumes logistic resource dynamics in eq. (23). A different closure, even one with the same qualitative idea of resource limitation, can produce different scaling exponents, which would break the proposed mapping from size-distribution slope to forest condition. The paper also ships no code and gives no empirical validation, so the closure question is not settled by numerical evidence. I therefore see no reason to change the reader's CONDITIONAL verdict; the concern is real and should be addressed before the central claim is accepted as a general mechanism. I would not reject the paper, because the framework is transparent, the stationary solution is derived, and the limitations are openly discussed; the appropriate resolution is to require an explicit mechanistic justification or sensitivity analysis for equations (5) and (6).","tokens_in":17644,"tokens_out":17464,"duration_ms":194376,"concrete_test":"Closure-sensitivity check: solve the stationary version of eq. (25) with the resource-mortality term replaced by alternative per-capita forms such as dR = c h^q (R/Ra)^p for q = 0, 1 and p = 1, 2, keeping eq. (5) unchanged, and compute the small-size exponent. If the exponent is no longer -R[ρ]/Ra, then the exponent readout in eq. (14) is an artifact of the specific closure chosen; if the exponent is unchanged, the concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The small-size exponent -a in eq. (14) is produced entirely by the linear mortality term \\tilde g(h) a/h from eq. (6), as it appears in the stationary equation (25). This term is not derived from metabolic scaling: the Methods section postulates logistic resource dynamics (eq. 23) and then chooses the mortality that reproduces it. A plausible alternative such as dR = c h^q (R/Ra)^p would change both the exponent and its interpretation. Likewise, the large-size exponent -(1+2H) is fixed by the h^{2H}ρ dependence in eq. (5). Thus the paper's central claim—that the observed size-distribution exponent is a mechanistic readout of forest condition—rests on two hand-written functional forms. The Discussion lists limitations of the model, but it does not address the closure dependence of the exponents; stability of ρ* is also asserted rather than proved, but the mortality closure is the more basic issue.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper develops a spatially explicit, size-structured model of forest dynamics based on a McKendrick–von Foerster equation, incorporating metabolic-scaling growth, Beer-Lambert light attenuation, two density-dependent mortality terms (space competition and resource competition), and nonlocal recruitment through seed dispersal. In the spatially homogeneous mean-field limit, the authors derive an exact stationary solution which, in the intermediate asymptotic regime, reduces to a dual power-law size distribution with exponents -a for small sizes and -(1+2H) for large sizes, separated by a crossover height h_c (eqs. 14-15). They interpret the exponent a, defined as the resource saturation ratio R[ρ*]/R_a, as a mechanistic readout of forest condition: a ≈ 1+2H indicates space limitation, while a ≪ 1+2H indicates resource limitation. The paper then analyzes the effect of a strip boundary on the stationary state, showing that small trees are disproportionately depleted while total resource consumption is barely affected, and extends this to two species with different dispersal lengths, finding that the long-dispersal species declines over very long timescales. The authors claim that these results connect individual-level metabolic scaling to community-level size-distribution scaling and provide a unifying dynamical perspective on forest scaling laws.","tokens_in":17950,"tokens_out":8831,"duration_ms":89029,"significance":"If the model's closures are accepted, the paper makes a valuable contribution: it provides an analytically tractable framework linking allometric growth, competition, and dispersal to an exact stationary size distribution, and the dual-power-law prediction with a crossover is precise and potentially falsifiable against forest inventory data. The boundary-effect result (edge suppression of small trees with weak impact on total resource consumption) and the dispersal-mediated competitive imbalance are interesting and non-obvious. The authors are transparent about the mean-field and homogeneity assumptions and provide numerical simulations for the spatial results. However, the significance is currently limited by the fact that the central 'exponent as forest condition' readout is largely built into the mortality closures rather than derived from metabolic scaling, and the predictive content of the exponent–condition relationship is not yet demonstrated against data beyond qualitative agreement with earlier disturbance studies.","major_comments":[{"comment":"The two mortality closures are not derived from metabolic scaling; the resource-consumption dynamics in eq. (23) is a postulated logistic form, and the mortality term d_R = \\tilde{g}_α(h)/h · R[ρ]/R_a in eq. (6) is the term that reproduces it. Similarly, the space-competition term d_S = d_α I_α h^{2H}ρ in eq. (5) is a direct assumption about the interaction range. The Discussion's statement that the scaling regimes 'arise directly from explicit demographic processes rather than from phenomenological assumptions' is therefore unsupported for the central exponents. A plausible alternative closure, for example d_R ∝ h^q (R/R_a)^p, would change both the predicted exponent and its interpretation. Please provide a mechanistic derivation of eqs. (5)-(6) from explicit biological processes, or a sensitivity analysis showing that the two-exponent structure and the crossover height are robust to plausible variations of these functional forms, or re-frame the claims as explicitly conditional on the closures.","section":"Methods 'Resource competition'; eqs. (5)-(6), (23)"},{"comment":"The small-size exponent is, by construction, the resource saturation ratio a = R[ρ*]/R_a that appears in the mortality term (6), and the statement that a ≪ 1+2H indicates resource limitation is therefore close to a definitional identity within the model. The predictive content of the readout lies in how a and the crossover height h_c depend on the environmental parameters (R_a, κ, γ, h_0, H, h_u), but the paper does not provide explicit expressions for a(R_a, κ, ...) or a quantitative comparison with measured size-distribution exponents, only a qualitative agreement with the disturbance literature. Please derive the parametric dependence of a and h_c, or confront the model's predicted exponent–crossover relationship with forest inventory data, so that the 'exponent as forest condition' claim becomes a falsifiable prediction.","section":"Results 'Exact stationary solution'; eqs. (14)-(15)"},{"comment":"The global stability of the stationary density ρ*(h) is asserted via an appeal to a 'formal time-dependent exact solution' and to numerical simulations, with no proof or precise conditions. Because the boundary and multispecies analyses all start from the premise that the system converges to ρ*(h), this gap is load-bearing. Please provide a proof of (or a precise statement of the conditions for) global convergence, or soften the claim to local stability for the parameter regimes used in the simulations.","section":"Results 'Exact stationary solution', stability paragraph"},{"comment":"The list of biological limitations (identical dynamics across species, homogeneous environment, isotropic dispersal, single effective resource) omits the most fundamental source of uncertainty for the central claim: the dependence of the predicted exponents and the crossover on the hand-written mortality closures in eqs. (5)-(6). Please add an explicit discussion of this limitation, and note that alternative closures would alter the exponents and their interpretation.","section":"Discussion, last paragraph"}],"minor_comments":[{"comment":"The sentence 'The critical height h_c depends on the the sapling recruitment rate' contains a duplicated definite article; remove 'the the'.","section":"Results 'Exact stationary solution'"},{"comment":"'supporting that it is globally stabile' should read 'globally stable'.","section":"Results 'Exact stationary solution', stability paragraph"},{"comment":"The hypergeometric argument '2F1(2,δδ+ 1;z)' appears to contain a typo; it should presumably be '2F1(2, δ, δ+1; z)' or the intended pair of parameters.","section":"Methods 'Stationary solution', eq. (29)"},{"comment":"The crown-radius scaling r_c ∼ h^H is cited to both [19] and [39]; [39] is a study of seedling-layer abundance patterns and does not appear to support this allometric relation. Please correct the citation.","section":"Introduction; references [19], [39]"},{"comment":"The decay timescale τ_ξ is estimated by an exponential fit in the range 50 ≤ τ ≤ 300, but the paper reports no confidence intervals or details of the fitting procedure. Please specify the fit, the error bars, and the sensitivity to the chosen time window.","section":"Results 'Interplay of boundary effects and inter-species interactions'; fig. 4C inset"}],"recommendation":"major_revision","confidential_remarks":"To the editor: The core concern, also noted by the reader, is that the central 'exponent as forest condition' readout hinges on the hand-written mortality closures in eqs. (5)-(6). The paper's Discussion overstates their mechanistic status, and this should be addressed either by derivation, robustness analysis, or a more careful framing. The manuscript is mathematically sound given the closures, and the exact stationary solution and boundary-effect results are valuable. Please also verify the citation for crown allometry (ref. [39]) and note that the paper leans on the authors' own recent arXiv preprints (refs. [38] and [74]) for background; these should be checked for publication status."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know before reading. First, this is a worthwhile theory paper: it solves a McKendrick–von Foerster equation with growth, shading, density-dependent mortality, and nonlocal recruitment, and shows that the stationary size distribution has two power-law regimes separated by a crossover set by seed injection and resource availability. The large-size exponent −(1+2H) matches the known metabolic scaling result, and the small-size exponent is the resource saturation ratio a. The boundary effect—edges suppress small trees and can slowly exclude longer-dispersal species—is new and plausible.\n\nSecond, the central claim that the size-distribution exponent is a mechanistic readout of forest condition is not as secure as the writing implies. The small-size exponent −a is produced entirely by the mortality closure in eq. (6), dR=(g/h)(R[ρ]/Ra), which is written by hand to make the resource dynamics logistic. The space-competition term is similarly postulated. The paper does not show these forms follow from metabolic scaling, and a different plausible closure, such as mortality proportional to h^q(R/Ra)^p, would change both exponents and their ecological interpretation. The authors present the forms clearly and analyze their consequences, but they do not test sensitivity to alternative closures or fit the model to data.\n\nThe mathematics is internally consistent; the stationary solution and its asymptotic reduction to eq. (14) check out. The self-consistent determination of a is implicit but legitimate. Stability is asserted rather than proved, although numerical simulations support it. The boundary results are numerical only, but the mechanism is intuitive. No code or data are shipped; that is acceptable for a theory paper but limits reproducibility of the figures.\n\nThis paper is for ecologists and physicists working on size-structured community models, and for anyone interested in connecting allometric scaling to forest structure. It deserves a serious referee. I would ask for (i) a sensitivity analysis of the exponents to the mortality closures, or a derivation of eq. (6) from a more mechanistic resource-competition model, and (ii) an explicit statement that the small-size exponent is model-dependent, not a universal prediction. If those are addressed, it is a solid contribution.","headline":"A tractable forest-size model with a novel dual power-law crossover and a boundary-driven dispersal-exclusion mechanism, but the 'exponent as forest condition' readout rests on hand-written mortality closures that need a sensitivity test.","tokens_in":18432,"tokens_out":2866,"would_cite":false,"duration_ms":30264,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["92D40","92D25"],"pacs":[],"model":"deepseek-v4-flash","headline":"A solvable forest model shows that the shape of a tree-size distribution is a mechanistic readout of whether a forest is limited by space or by resources.","keywords":["tree-size distribution","metabolic scaling","forest dynamics","power-law scaling","size-structured population model","dispersal kernel","resource competition","space competition"],"falsifier":"Measure, in a set of forest plots, the small-size and large-size slopes of the size distribution together with independent estimates of the resource saturation ratio $a=R/R_a$ and the crown allometry exponent $H$. If the small-size slope does not track $a$, or the crossover height does not shift with measured recruitment and $R_a$ as eq. (15) predicts, the mortality closure is falsified. A simpler check: in a plot where fertilization or irrigation raises $R_a$, the model predicts the crossover $h_c$ moves downward; if it stays fixed while both slopes change, the proposed mechanism is wrong.","tokens_in":1689,"feed_emoji":"🌳","tokens_out":2018,"duration_ms":76421,"temperature":0.7,"pith_summary":"This paper argues that the shape of a forest's tree-size distribution is not merely a summary statistic but a mechanistically readable record of what limits tree growth. Starting from metabolic energy balance for individual trees, the authors build a size-structured dynamical model with light shading, density-dependent mortality, and seed dispersal, and solve it exactly in the spatially homogeneous limit. The stationary solution has two power-law regimes: small trees follow a slope set by resource saturation, while larger trees follow a slope set by crown allometry and spatial competition. The height at which the distribution switches between these regimes is controlled by seed recruitment and local resource availability. If the model is right, a forest inventory can indicate whether resource limitation or space competition dominates, and how far the forest sits from an undisturbed reference state.","feed_headline":"Forest size data can reveal whether space or resources limit trees","feed_subtitle":"A solvable forest model shows the size-spectrum slope shifts with resource saturation, linking structure to condition.","key_machinery":"The machinery is a McKendrick–von Foerster continuity equation in height space for tree density $\\rho_\\alpha(x,h,t)$, coupled to three process terms: ontogenetic growth derived from metabolic energy balance (growth velocity $g_\\alpha(h)=g_0(1-h/h_u)$), Beer–Lambert light attenuation by taller neighbors, and a mortality rate split into space competition $d_S=d\\,I\\,h^{2H}\\rho$ and resource competition $d_R=(g(h)/h)\\,R[\\rho]/R_a$, where $R[\\rho]=\\int h^{1+2H}\\rho\\,dh$ is total community metabolic demand. Recruitment enters as a nonlocal boundary condition at seedling height through a dispersal kernel. In the homogeneous limit the stationary equation reduces to a solvable form whose asymptotic expansion gives the two power laws and the crossover of eqs. (14)–(15); the same machinery, with a Bessel-function kernel for a strip geometry, produces the boundary and multispecies results.","core_discovery":"The central discovery is that the stationary tree-size density $\\rho^*(h)$ obeys a closed-form two-term expression (eq. 14) in which a resource-competition term produces a small-size power law $\\rho^*(h)\\sim h^{-a}$ and a space-competition term produces a large-size power law $\\rho^*(h)\\sim h^{-(1+2H)}$. Here $a=R[\\rho^*]/R_a$ is the local resource saturation ratio and $H$ is the crown allometry exponent, so the crossover height $h_c$ where the two terms balance is set by recruitment intensity and available resources. The paper claims this makes the observable size-distribution slope a mechanistic readout of forest condition: $a\\approx 1+2H$ signals space-limited dynamics, while $a\\ll 1+2H$ signals resource-limited dynamics. It further claims that in spatially explicit settings, boundary disturbances suppress small trees much more than large ones, and that in two-species systems a longer dispersal range becomes a systematic disadvantage near boundaries, driving abundance declines that extend far beyond the dispersal length.","pith_inferences":["A direct test would use permanent-plot inventories with simultaneous light and resource measurements to estimate $a$ and $H$ independently, then compare the predicted versus observed exponent and crossover height.","The same two-regime structure may apply to other sessile, size-structured communities such as coral, seagrass, or mussel beds, where canopy shading and space occupancy play analogous roles.","The model suggests a null expectation that disturbance gradients should shift the small-size exponent before the large-size exponent, a pattern that remote sensing of canopy height distributions could search for.","Because the homogeneous solution is independent of dispersal kernel shape, persistent deviations from the predicted dual power law in real forests could serve as a diagnostic for spatial heterogeneity or boundary effects."],"forward_implications":["A forest whose size spectrum follows a single power law with slope close to $-(1+2H)$ is operating in the space-limited, resource-abundant regime, while a visibly smaller slope indicates resource saturation with $a<1+2H$.","The crossover height $h_c$ moves downward when sapling recruitment is strong or available resources are high, so surveys that resolve small trees can locate the crossover and infer recruitment and resource status.","Edge and fragmentation effects should appear mainly in the small-tree sector, so substantial declines in total density can occur even while total resource consumption stays nearly uniform.","In mixed forests, species with longer dispersal ranges should decline near forest edges and possibly across the whole fragment, with the exclusion timescale growing as dispersal ranges converge.","In a homogeneous environment, dispersal range alone does not change stationary relative abundances, so observed dispersal-related differences in abundance imply spatial structure or disturbance."],"supporting_citations":[{"why":"Supplies the metabolic energy-balance and allometric relations from which the individual growth law $g(h)$ is derived.","marker":"[9–13, 17]"},{"why":"Provides the crown-radius scaling with height, $r_c\\sim h^H$, and the self-similar forest-size scaling that sets the large-size exponent $1+2H$.","marker":"[19]"},{"why":"Gives the Beer–Lambert canopy light attenuation used in the shading factor $I(x,h,t)$.","marker":"[40, 41]"},{"why":"Provides the McKendrick–von Foerster size-structured population formalism that the dynamical equation extends.","marker":"[23, 26, 31]"},{"why":"Underlies the neutral-community assumption that lets the multispecies dynamics be written for total density.","marker":"[42]"},{"why":"Empirical links between size-distribution exponents and disturbance that the model interprets mechanistically.","marker":"[49–51]"},{"why":"Frames seed-dispersal trade-offs and spatial recruitment patterns motivating the nonlocal boundary condition.","marker":"[35]"}],"fun_headline_variants":["Tree-size law shows when space versus resources rule forests","Forest dynamics distilled into a two-part power law","How tree size distribution reveals forest competition type","Boundary effects can ripple through forests far beyond seeds","A solvable model links tree sizes to resource saturation"],"cache_read_input_tokens":20608,"weakest_assumption_plain":"The load-bearing premise is that mortality from space competition scales as $d\\,I\\,h^{2H}\\rho$ and mortality from resource competition as $(g(h)/h)\\,R[\\rho]/R_a$; these two functional forms are postulated rather than derived, and they are exactly what produces the two power-law exponents and the crossover.","fun_headline_variants_meta":{"raw":{"variants":["Tree-size law shows when space versus resources rule forests","Forest dynamics distilled into a two-part power law","How tree size distribution reveals forest competition type","Boundary effects can ripple through forests far beyond seeds","A solvable model links tree sizes to resource saturation"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000775,"raw_usage":{"total_tokens":3434,"prompt_tokens":958,"completion_tokens":2476,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":574,"completion_tokens_details":{"reasoning_tokens":2403}},"tokens_in":574,"tokens_out":2476,"duration_ms":17414,"temperature":1.0,"reasoning_tokens":2403,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T00:22:44.950805+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure, in a set of forest plots, the small-size and large-size slopes of the size distribution together with independent estimates of the resource saturation ratio $a=R/R_a$ and the crown allometry exponent $H$. If the small-size slope does not track $a$, or the crossover height does not shift with measured recruitment and $R_a$ as eq. (15) predicts, the mortality closure is falsified. A simpler check: in a plot where fertilization or irrigation raises $R_a$, the model predicts the crossover $h_c$ moves downward; if it stays fixed while both slopes change, the proposed mechanism is wrong.","supporting_citations":[{"cited_title":"Simini, T","cited_arxiv_id":null,"evidence_quote":"Provides the crown-radius scaling with height, $r_c\\sim h^H$, and the self-similar forest-size scaling that sets the large-size exponent $1+2H$."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Underlies the neutral-community assumption that lets the multispecies dynamics be written for total density."},{"cited_title":"Nathan and H","cited_arxiv_id":null,"evidence_quote":"Frames seed-dispersal trade-offs and spatial recruitment patterns motivating the nonlocal boundary condition."}],"review_version":1}