{"id":"0c306ac2-b2eb-4cb4-829a-e8ce4dbe7fa9","arxiv_id":"2411.18792","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"A multistage spatial model identifies the critical release profile needed for Wolbachia to persist and spread, and simulates how mitigation, habitat, and seasonality change that threshold.","lead":"The authors build an 8-equation reaction-diffusion model of how Wolbachia infections spread through Aedes aegypti mosquitoes, adding eggs, larvae, males and females to earlier spatial models. They compute the minimum 'critical bubble' release profile and use simulations to compare release strategies, finding that adult-stage pre-release control, dry-region releases, and pre-wet-season timing each lower the required release size.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Gaussian diffusion assumption is load-bearing: fat-tailed or sex-dependent dispersal could reverse the central strategy rankings, yet this is calibrated from a single mark-release-recapture study.","rationale":"The reader identified the same weakest assumption: the Gaussian diffusion with a single D for all adults. I agree this is the most load-bearing concern. The central claim is not merely that a threshold exists—which is a general property of bistable systems—but that specific intervention rankings hold (adult mitigation > aquatic mitigation, dry release lowers threshold, pre-wet-season release lowers threshold). These rankings emerge from the spatial dynamics; in particular, the adult-vs-aquatic mitigation ranking is explicitly contrasted with the ODE model [34], where the two are comparable. Therefore, if the diffusion representation is mis-specified, the rankings could be artifacts. The sensitivity analysis shows diffusion strongly controls bubble width and wave width, and the authors themselves cite evidence that fat-tailed kernels alter wave velocity. The calibration from a single Australian town with a Gaussian conversion is thin. A concrete test with a fat-tailed kernel would directly assess whether the qualitative conclusions survive. I did not elevate the threshold algorithm to the top concern because, despite the heuristic phrasing, the algorithm's output was cross-validated against the 2-PDE model (Fig 3B) and the sign-change logic is plausible; the diffusion assumption has no such independent validation. The verdict remains CONDITIONAL because the paper's qualitative insights are valuable but require robustness checks on dispersal.","tokens_in":20989,"tokens_out":8551,"duration_ms":75700,"concrete_test":"Re-run the three headline scenario comparisons (adult vs aquatic pre-release mitigation in Section 3.2, dry vs wet region release in Section 3.3, and seasonal release timing in Section 3.4) using a fat-tailed dispersal kernel calibrated to the same mark-release-recapture data (e.g., the kernel from Turelli & Barton 2017) instead of the Gaussian kernel, while keeping all other parameters fixed. If the qualitative ranking 'adult mitigation reduces threshold more than aquatic mitigation' persists under the fat-tailed kernel, the diffusion concern is mitigated; if the ranking reverses or becomes comparable to the ODE result, the central strategy recommendations are not robust to the dispersal assumption and the paper's conclusions require significant qualification.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 2.1 assumes isotropic Gaussian diffusion with identical coefficient D=200 m2/day for both sexes and all adult cohorts, calibrated in Section 5.2 from the Pentland mark-release-recapture mean distance traveled (25 m/day) via the Gaussian mean absolute deviation relation E1=sqrt(pi D). This single assumption underpins the paper's most novel qualitative results: (1) the claim that adult-stage pre-release mitigation reduces the threshold release number more than aquatic-stage mitigation (Section 3.2) is explicitly attributed to spatial diffusion, because the companion ODE model in [34] showed comparable impacts; (2) the critical bubble width and wave width are dominated by D in the sensitivity analysis (Fig 4C,E); and (3) wave propagation and dry/wet interface behavior depend on diffusion. The authors acknowledge in Section 4 that non-Gaussian dispersal kernels can substantially change wave velocity and that some studies report sex-dependent dispersal, but they do not test whether these alternative assumptions alter their qualitative conclusions. If the true dispersal kernel is fat-tailed or D is much smaller (e.g., D=50), the spatial effects that drive the adult-vs-aquatic mitigation ranking could vanish or reverse, directly undermining a headline recommendation. Since the central claim of the paper is not just the existence of a threshold but the strategy rankings derived from the spatial model, this unvalidated modeling choice is the most load-bearing concern.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops an 8-component reaction-diffusion PDE model for Wolbachia spread in Aedes aegypti, tracking eggs, larvae/pupae, males, and females in infected and uninfected cohorts, with diffusion acting on adult stages only. The authors numerically identify a 'critical bubble'—the threshold release profile for self-sustaining infection—and then use the model to compare pre-release mitigation strategies (adult vs aquatic), habitat modification, release in dry vs wet regions, and seasonal timing. The manuscript is carefully parameterized from laboratory and field data and includes a global sensitivity analysis via LHS/PRCC, as well as a comparison with the earlier 2-PDE model of the authors.","tokens_in":21270,"tokens_out":7008,"duration_ms":59095,"significance":"The paper's contribution is a spatially explicit, stage-structured model that bridges ODE threshold theory and 2-D release strategies for Wolbachia-based population replacement. If the results are robust, the findings on adult-targeting mitigation and seasonal timing could inform actual release programs. Strengths include transparent parameter tables, a sensitivity analysis, and explicit acknowledgment of modeling limitations. However, the central quantitative claims rest on a heuristic threshold definition whose convergence is not documented, and on a single mark-release-recapture study for the diffusion coefficient under a Gaussian assumption—a parameter that the paper's own sensitivity analysis shows to be highly influential.","major_comments":[{"comment":"The threshold 'critical bubble' is identified by the root-finding function J(c) defined in Eq. (3), but the text states that 'J(c)=0 has no exact root' and that the critical bubble is determined from the 'plateau in the time series of Fig 2C around t=1000'. This is a heuristic; the manuscript does not specify the plateau-detection tolerance, the final time T, or the time step Δt used in the reported thresholds, nor does it provide a convergence study. Because all subsequent results (e.g., the threshold release numbers X≈6.6/m² in Section 3.3 and the mitigation comparisons in Fig. 5) are computed relative to this c*, an ill-defined threshold could alter the qualitative rankings. Please provide a precise algorithmic definition, a convergence check in T and Δt, and an uncertainty estimate for c*.","section":"Section 2.2.2 (Eqs. 3-4)"},{"comment":"The baseline diffusion coefficient D=200 m²/day is derived from one mark-release-recapture study (Pentland) via the Gaussian mean-absolute-deviation relation E1=√(πD)=25, and the model assumes isotropic Gaussian diffusion with equal D for both sexes and all adult life stages. This assumption is load-bearing: Fig. 4C,E shows that D dominates the bubble and wave widths, and Section 3.2.2 explicitly attributes the ranking 'adult-stage mitigation better than aquatic-stage mitigation' to 'spatial diffusion dynamics among adults'. The Discussion admits that fat-tailed kernels and sex-dependent dispersal can change wave velocity, but no alternative-kernel or alternative-D simulations are presented. The paper should test whether the strategy rankings in Sections 3.2-3.4 survive (i) non-Gaussian kernels, (ii) sex-specific diffusion coefficients, and (iii) D values near the ends of the Table 1 range (100 and 300 m²/day).","section":"Section 5.2 and Section 3.2.2"},{"comment":"Table 1 states the diffusion coefficient range as 100-300 m²/day, but the sensitivity analysis in Table B.1 restricts D to 150-250 m²/day (a ±25% band around the baseline). Since the global SA identifies D as a dominant parameter for the width of the critical bubble and the infection wave, the PRCC results do not cover the full stated uncertainty. The analysis should be repeated over the full range (or the narrower range must be justified), and the qualitative conclusions should be checked for robustness.","section":"Table 1 vs. Table B.1"}],"minor_comments":[{"comment":"The fitted rainfall functions contain ambiguous notation, e.g., '0.03((1.47 sin(2π(t/365 + 0.17)) + 1.47)2.54 + 1)' and '−0.34 sin(2π1.88(t/365 + 0.08)) + 25.93'; please clarify the intended exponents and frequencies (the superscript '2.54' appears to be lost).","section":"Table 2"},{"comment":"The statement that all codes will be made available on GitHub upon acceptance is helpful, but for a numerical study, providing the code at the time of review would greatly aid verification of the threshold algorithm and the reported results.","section":"Section 2.3"},{"comment":"The 2-D simulation setup, including domain size, grid spacing, and boundary condition implementation, is not fully described; specifying these would improve reproducibility.","section":"Section 2.3"},{"comment":"The heatmaps in panels A, D, E, G, and H are difficult to read quantitatively; consider adding contour labels or a table of threshold release numbers for the headline scenarios.","section":"Fig 5"},{"comment":"The derivation of D from the Cairns weekly MDT values uses E_Δt = √(πD Δt) but the text should explicitly state that Δt is measured in days; additionally, the choice to prefer the Pentland daily estimate over the Cairns estimates (D≈120-570) could use a brief justification beyond 'daily dispersal gives a more reliable characterization'.","section":"Section 5.2"}],"recommendation":"major_revision","confidential_remarks":"The paper is a useful extension of the authors' prior ODE work [34] and the 2-PDE model [23], but the novelty is incremental and the central quantitative results rest on a non-rigorous threshold heuristic and a single-study diffusion parameter. The revision should focus on robustness checks (alternative kernels, D ranges, and threshold-algorithm convergence) rather than new biological claims. The code availability statement should be fulfilled at revision, as numerical verification is otherwise difficult."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper does something real: it builds an 8-PDE reaction-diffusion model with explicit egg, larval, male, and female stages, and uses it to ask questions the earlier 2-PDE model could not, like larviciding versus adult spraying, habitat modification, and dry-versus-wet release sites. The parameterization is careful, drawing on lab and field studies, and the two seasonal case studies (Cairns and Yogyakarta) are a nice step toward application. The sensitivity analysis is informative, and the finding that adult-stage pre-release mitigation beats aquatic-stage mitigation is interesting and mechanistically plausible: spatial diffusion lets released infected adults fill the gaps created by adult removal, which an ODE model would miss. The critical-bubble framework is inherited from the authors' prior work, but applying it to this fuller model is a legitimate extension, and the appendix shows the threshold is insensitive to release shape. Credit where due: this is a competently executed modeling study that gives concrete, testable guidance for release programs.\n\nThe soft spots are real but, in my view, not disqualifying. The biggest is the dispersal assumption: isotropic Gaussian diffusion with the same D for both sexes, calibrated from one mark-release-recapture study in Pentland. The sensitivity analysis shows D strongly controls bubble width and wave width, and the adult-versus-aquatic ranking is explicitly attributed to spatial diffusion. The authors mention in the Discussion that fat-tailed kernels and sex-dependent dispersal exist and could change wave velocity, but they do not test whether their qualitative rankings survive those alternatives. That is a genuine gap, though it is an acknowledged limitation of a modeling paper, not a hidden flaw. A second issue is that the threshold is identified by a numerical heuristic (the J(c)=0 root finder that has no exact root, with the critical bubble inferred from a plateau). The authors are transparent about this, and the release-shape robustness test helps, but it is not a rigorous bifurcation computation. Third, the code is not yet available, so I could not verify the numerical results; the promise of GitHub upon acceptance is good but should be enforced. The seasonal rescaling of temperature-dependent traits is also fairly hand-calibrated, though again transparently described.\n\nWho is this for? Researchers working on Wolbachia release logistics or spatial mosquito models. It is a useful qualitative guide, not a quantitatively validated prediction. The central claims are internally consistent and the limitations are mostly stated, so I would take the paper seriously. My recommendation: send it to peer review. Require the code/data as a condition of acceptance, and ask the authors to add a robustness section showing whether the adult-versus-aquatic ranking and the dry-release recommendation survive fat-tailed or sex-dependent dispersal kernels and a lower D. That would turn a good qualitative study into a more durable one.","headline":"A solid, genuinely useful extension of the authors' spatial Wolbachia work; the new 8-PDE model supports qualitative strategy rankings, but the single-Gaussian-diffusion assumption deserves a robustness check before the adult-over-aquatic recommendation is taken as robust.","tokens_in":21795,"tokens_out":1757,"would_cite":true,"duration_ms":19594,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["92D25","92D30","35K57"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper claims that Wolbachia establishment in a two-dimensional mosquito population is governed by a spatial critical-bubble threshold, with infection waves propagating only beyond it.","keywords":["Wolbachia","Aedes aegypti","reaction-diffusion model","critical bubble","population replacement","release strategy","seasonality"],"falsifier":"A field measurement of the daily mosquito displacement distribution at a candidate release site would test the Gaussian-equal-diffusion kernel; if the measured kernel is fat-tailed or the sexes differ in mean distance, the predicted critical bubble and wave speed would not match the model. A controlled release just above the predicted 60%-at-center threshold in a site with larval carrying capacity near one per square meter should show the infection fraction rising to the endemic steady state; if it decays instead, the critical-bubble claim is contradicted.","tokens_in":20758,"feed_emoji":"🦟","tokens_out":6487,"duration_ms":60172,"temperature":0.7,"pith_summary":"This paper argues that whether a local release of Wolbachia-infected mosquitoes establishes a self-sustaining infection is governed by a spatial threshold it calls the critical bubble: a bubble-shaped profile of infection fraction that must be exceeded near the release center. Above that bubble the infection spreads as a traveling wave through the two-dimensional landscape; below it the infection collapses. The paper builds this conclusion from an eight-equation reaction-diffusion model that tracks eggs, larvae and pupae, males, and females for infected and uninfected mosquitoes, with adult dispersal modeled as diffusion. It then uses the model to rank release strategies, concluding that adult-stage pre-release mitigation, dry-region release, and release just before the wet season each lower the required release number, while reducing breeding sites can raise it.","feed_headline":"A 'critical bubble' decides if Wolbachia release succeeds","feed_subtitle":"Adult-targeted sprays and pre-wet-season timing lower the Wolbachia release threshold, an eight-stage spatial model shows.","key_machinery":"The machinery is an 8-PDE reaction-diffusion system with state variables for eggs, larvae and pupae, males, and females, each split into uninfected and Wolbachia-infected cohorts. Maternal transmission and cytoplasmic incompatibility couple the cohorts: uninfected females mating with infected males produce no viable offspring, while infected females pass the infection to a fraction vw of their eggs. Adult males and females move by isotropic diffusion with coefficient D = 200 $m^{2}$/day calibrated to mark-release-recapture data; aquatic stages do not move. The critical bubble is computed by a root-finding functional J(c) that detects, for each candidate initial release level c, whether the infection fraction at the release center rises to the endemic steady state, collapses to zero, or passes through a stagnant bifurcation plateau. That functional pins the threshold release profile.","core_discovery":"The central claim is that in a spatially explicit, life-stage-resolved model of Aedes aegypti with wMel Wolbachia, the invasion threshold is not a single infection fraction but a spatially varying critical bubble. At baseline parameterization the model finds a threshold of about 60% infection at the release center, with the required fraction falling with distance from the release center. Releasing beyond the bubble produces a self-sustaining infection that propagates outward as a wave, while releasing below it lets the infection die out. The paper also claims that this model changes practical guidance: adult-stage pre-release mitigation reduces the threshold release number much more than aquatic-stage mitigation, reducing breeding sites can increase the threshold, releases in dry areas lower the threshold (though waves may stall at dry-wet interfaces), and release timing just before the wet season minimizes the threshold.","pith_inferences":["Beyond the paper, measuring the actual daily displacement distribution at a candidate release site would directly test the Gaussian-equal-diffusion assumption; if the kernel is fat-tailed or males and females differ in flight distance, the critical bubble and strategy rankings could shift.","Beyond the paper, the model implies a potential tension between suppression and replacement strategies: habitat modifications that reduce wild mosquito abundance by removing breeding sites may make Wolbachia establishment harder, so programs should fix their control objective before choosing pre-release actions.","Beyond the paper, because maternal transmission rate dominates every threshold and wave metric, field monitoring that detects even small reductions in transmission below 100% would materially change release requirements; this is a prioritized parameter to measure.","Beyond the paper, the same modeling framework could be parameterized for other Wolbachia strains or for Anopheles mosquitoes, where fitness costs and cytoplasmic incompatibility strength differ, to check whether the qualitative rankings of release strategies persist."],"forward_implications":["A local release whose infection fraction sits below the critical bubble at the release center dies out even if the same fraction would persist in a spatially uniform model; release programs should be designed around the bubble profile, not a single threshold fraction.","Pre-release spraying that removes adult mosquitoes lowers the required release number substantially more than larviciding, with a consistent optimal release radius near 200 meters under baseline parameters.","Habitat modification that removes breeding sites and lowers larval carrying capacity can raise the threshold for Wolbachia establishment, in some parameter ranges making establishment essentially unachievable.","Releasing in the drier side of a dry-wet landscape lowers the release threshold, but the infection wave can slow or stop at the interface when carrying capacity contrast is large, for instance at a 1:5 ratio.","Seasonal timing matters: releasing just before the wet season, when wild mosquito abundance is low, minimizes the required release threshold in both tropical case-study sites."],"supporting_citations":[{"why":"Supplies the earlier 2-PDE spatial model whose numerical algorithm for identifying the critical bubble is adapted here and whose threshold is compared against the 8-PDE result.","marker":"[23]"},{"why":"Provides the prior ODE model that established the bistable threshold condition and the basic reproduction number used to frame Wolbachia invasion.","marker":"[22]"},{"why":"Provides the prior ODE model of sustained Wolbachia transmission that yields the endemic steady state and threshold formulas used in the algorithm and sensitivity analysis.","marker":"[34]"},{"why":"Introduces the reaction-diffusion framework for bistable spatial waves of Wolbachia that motivates modeling infection spread as a traveling wave.","marker":"[30]"},{"why":"Models Wolbachia spread with non-Gaussian dispersal kernels and supplies context for the diffusion parameter range and for caveats about wave velocity.","marker":"[26]"},{"why":"The Pentland mark-release-recapture study whose 25 m/day mean distance traveled calibrates the baseline diffusion coefficient D = 200 m^2/day.","marker":"[43]"},{"why":"The Cairns mark-release-recapture study used to set the diffusion coefficient range and to illustrate environmental variability in mosquito dispersal.","marker":"[44]"},{"why":"Supplies wMel cage-experiment values for fitness costs, maternal transmission, cytoplasmic incompatibility, and adult lifespan used in parameterization.","marker":"[4]"},{"why":"Provides field estimates of egg hatch rates, fecundity, and Wolbachia transmission used to set egg mortality and fecundity parameters.","marker":"[11]"}],"fun_headline_variants":["Spatial critical bubble sets Wolbachia release threshold","Critical bubble size predicts Wolbachia invasion success","Dry-season release lowers Wolbachia threshold","Adult-targeted sprays reduce Wolbachia release threshold"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The model assumes adult mosquitoes spread by simple random diffusion with the same average step size for males and females, a step size estimated from marked mosquitoes in one Australian town; if real movement is long-distance-jumpy or differs between sexes, the critical bubble and strategy advice could change.","fun_headline_variants_meta":{"raw":{"variants":["Spatial critical bubble sets Wolbachia release threshold","Critical bubble size predicts Wolbachia invasion success","Dry-season release lowers Wolbachia threshold","Adult-targeted sprays reduce Wolbachia release threshold"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001021,"raw_usage":{"total_tokens":4337,"prompt_tokens":1005,"completion_tokens":3332,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":621,"completion_tokens_details":{"reasoning_tokens":3271}},"tokens_in":621,"tokens_out":3332,"duration_ms":23761,"temperature":1.0,"reasoning_tokens":3271,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T10:53:17.146599+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A field measurement of the daily mosquito displacement distribution at a candidate release site would test the Gaussian-equal-diffusion kernel; if the measured kernel is fat-tailed or the sexes differ in mean distance, the predicted critical bubble and wave speed would not match the model. A controlled release just above the predicted 60%-at-center threshold in a site with larval carrying capacity near one per square meter should show the infection fraction rising to the endemic steady state; if it decays instead, the critical-bubble claim is contradicted.","supporting_citations":[{"cited_title":"Modeling Spatial Waves of Wolbachia Invasion for Controlling Mosquito-Borne Diseases","cited_arxiv_id":null,"evidence_quote":"Supplies the earlier 2-PDE spatial model whose numerical algorithm for identifying the critical bubble is adapted here and whose threshold is compared against the 8-PDE result."},{"cited_title":"Modeling the Transmission of Wolbachia in Mosquitoes for Controlling Mosquito-Borne Diseases","cited_arxiv_id":null,"evidence_quote":"Provides the prior ODE model that established the bistable threshold condition and the basic reproduction number used to frame Wolbachia invasion."},{"cited_title":"Modeling Sustained Transmission of Wolbachia among Anopheles Mosquitoes: Implications for Malaria Control in Haiti","cited_arxiv_id":null,"evidence_quote":"Provides the prior ODE model of sustained Wolbachia transmission that yields the endemic steady state and threshold formulas used in the algorithm and sensitivity analysis."},{"cited_title":"Deploying Dengue-Suppressing Wolbachia : Robust Models Predict Slow but Effective Spatial Spread in Aedes Aegypti","cited_arxiv_id":null,"evidence_quote":"Models Wolbachia spread with non-Gaussian dispersal kernels and supplies context for the diffusion parameter range and for caveats about wave velocity."},{"cited_title":"Aedes Aegypti Survival and Dispersal Estimated by Mark-Release-Recapture in Northern Australia","cited_arxiv_id":null,"evidence_quote":"The Pentland mark-release-recapture study whose 25 m/day mean distance traveled calibrates the baseline diffusion coefficient D = 200 m^2/day."},{"cited_title":"Mark-Release-Recapture Study to Measure Dispersal of the Mosquito Aedes Aegypti in Cairns, Queensland, Australia","cited_arxiv_id":null,"evidence_quote":"The Cairns mark-release-recapture study used to set the diffusion coefficient range and to illustrate environmental variability in mosquito dispersal."},{"cited_title":"The wMel Wolbachia Strain Blocks Dengue and Invades Caged Aedes Aegypti Populations","cited_arxiv_id":null,"evidence_quote":"Supplies wMel cage-experiment values for fitness costs, maternal transmission, cytoplasmic incompatibility, and adult lifespan used in parameterization."},{"cited_title":"Stability of the wMel Wolbachia Infection Following Invasion into Aedes Aegypti Populations","cited_arxiv_id":null,"evidence_quote":"Provides field estimates of egg hatch rates, fecundity, and Wolbachia transmission used to set egg mortality and fecundity parameters."}],"review_version":1}