Pith. sign in

REVIEW 3 major objections 4 minor 53 references

Mechanistic models for panel data: Analysis of ecological experiments with four interacting species

T0 review · 3 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read A stochastic juvenile-stage model with latent food and parasite compartments reproduces a four-species Daphnia panel better than the earlier deterministic sexual-switch model, and points to resource depletion as the driver of collapse.

desk verdict A solid PanelPOMP case study whose headline 'better than any alternative' is overstated because the Searle benchmark is scored from published predictions rather than refit under the same likelihood, though the core inference workflow is sound. read the letter →

arxiv 2506.04508 v2 pith:V3DE23LU submitted 2025-06-04 stat.AP q-bio.PEstat.CO

Bo Yang , Jesse Wheeler , Meghan A. Duffy , Aaron A. King , Edward L. Ionides This is my paper
classification stat.APq-bio.PEstat.CO MSC 62M0592D40
keywords PanelIteratedFilteringSIRJPFmodelPOMPDaphniapartiallyobservedMarkovprocesslatentfoodcompartmenthost-parasitedynamicsmesocosmexperiment
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tries to show that a stochastic mechanistic model with an explicit juvenile life stage and unmeasured food and parasite compartments can explain a mesocosm panel of two Daphnia species, their algal food, and a fungal parasite better than the earlier deterministic model that attributed population collapse to a switch to sexual reproduction. The payoff of establishing this is twofold: it gives ecologists a likelihood-based route to test rival mechanistic explanations of multispecies time series, and it offers a concrete reinterpretation of this particular experiment, namely that food depletion, rather than ephippia production, drives the observed Daphnia decline. A sympathetic reader would summarize the claim as: if this model is right, resource limitation plus age structure is sufficient to reproduce the data, and the unobserved alga and parasite dynamics are learnable from panel records of host densities alone.

What carries the argument

The load-bearing object is the latent food compartment $F_u(t)$ in Eq. (5), coupled to a latent parasite compartment $P_u(t)$ and an explicit juvenile stage $J_u^k(t)$; the estimation machinery is panel iterated filtering (PIF), a plug-and-play particle-filter algorithm that maximizes the likelihood for panel partially observed Markov process (POMP) models by perturbing particle parameters and resampling. The food compartment sets up competition between native and invasive Daphnia through consumption, the juvenile stage adds a maturation delay and food competition that the adult-only model lacks, and the parasite compartment mediates infection. PIF lets the authors maximize a likelihood that would otherwise be intractable, making AIC comparison, profile likelihoods, and Monte Carlo adjusted profile confidence intervals possible for a nonlinear, non-Gaussian panel model with shared parameters across replicated mesocosms.

What would settle it

Measure algal cell density (and ideally parasite spore density) on the same five-day schedule in replicate mesocosms of the four-species treatment, and compare the observed trajectory with the model's predicted latent $F_u(t)$ trajectory; if the measured algae do not crash sharply before the Daphnia collapse and rebound afterward, the central mechanism is refuted.

Watch

Extended reading notes

Core claim

The paper's central claim is that its SIRJPF2 model—a system of Itô stochastic differential equations tracking susceptible, infected, and juvenile Daphnia of both species, plus latent alga food $F_u(t)$ and parasite spore density $P_u(t)$—quantitatively describes the observed panel data better than any alternative considered. Maximum-likelihood fits place SIRJPF2 at log-likelihood $-880.56$ (AIC 1809.12), far ahead of the deterministic sexual-switch model of Searle et al. (2016a) at $-2483.49$ (AIC 5004.97) and ahead of negative-binomial benchmark regressions. Despite never being fitted to juvenile counts, the fitted model reproduces the held-out juvenile data, and its predicted latent trajectories show algae crashing and then rebounding as Daphnia populations first collapse and then recover. On the paper's own terms, this is evidence that resource depletion, mediated by a juvenile stage that competes for food and delays adult recruitment, is the operative mechanism rather than a reproductive switch to ephippia. The authors explicitly caution that the food mechanism remains an unvalidated assumption because food was not measured.

Load-bearing premise

The latent food compartment $F_u(t)$ in Eq. (5) is assumed to be a valid description of unmeasured algal dynamics; if it is mis-specified, the resource-depletion interpretation could be an artifact even though the statistical fit holds.

Editorial extensions

If this is right

  • If the latent food model is accepted, the Daphnia collapse in this experiment is attributable to algal resource depletion rather than to the sexual-switch and ephippia mechanism.
  • Age structure matters even when juveniles are rarely diseased: the juvenile stage improves statistical fit and reproduces the out-of-sample juvenile counts.
  • Stochastic dynamics suffice to explain differences between replicate mesocosms, since unit-specific parameters did not improve the AIC.
  • AIC-based comparison of mechanistic and non-mechanistic benchmark models is feasible for panel ecological data using PIF and marginalized panel iterated filtering.
  • The model's latent trajectories give testable predictions for the unmeasured alga and parasite densities.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • An experiment that directly measures algal cell density on the same five-day schedule, or manipulates food supply, could confirm or refute the resource-depletion mechanism; the model predicts a sharp algal crash followed by rebound as Daphnia decline.
  • The same panel POMP pipeline could be applied to other multi-species mesocosm or field panels, treating food and parasite states as latent whenever they are not measured.
  • Adding the male and ephippial female counts that the original experiment collected but neither model used would let the sexual-switch hypothesis be tested within the same likelihood framework.
  • The weak identifiability of the invasive-species birth, filtration, and infectivity parameters suggests that simplified host-only treatments are needed to separate those products, as the paper's supplementary comparisons begin to show.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The paper develops a PanelPOMP analysis of a Daphnia mesocosm experiment involving two Daphnia species, an algal food resource, and a fungal parasite. It proposes a stochastic differential equation model (SIRJPF2) with a latent food compartment, a latent parasite compartment, and an explicit juvenile stage, fits the model to replicated panel data using panel iterated filtering with a marginalized variant (MPIF), and compares it against alternative mechanistic models and negative-binomial benchmarks using AIC, Monte Carlo adjusted profile (MCAP) confidence intervals, and out-of-sample juvenile data. The authors report that SIRJPF2 fits the observed data better than all alternatives and use the fitted model to argue that resource depletion, rather than the sexual-switch mechanism of Searle et al. (2016), explains the Daphnia decline.

Significance. The methodological demonstration is valuable: it provides a practical plug-and-play, likelihood-based workflow for nonlinear non-Gaussian PanelPOMP models with shared and unit-specific parameters, and it ships code and data in a public repository. The out-of-sample juvenile validation, simulation bands, and MCAP intervals are genuine strengths. If the comparison issues raised below are addressed, the paper would be an instructive case study and a plausible challenge to the previous deterministic interpretation of this experiment.

major comments (3)
  1. [Table 2 and Section 5] The SIRJPF2 row in Table 2 reports 26 parameters, log-likelihood -880.56, and AIC 1809.12, but 2*26 - 2*(-880.56) = 1813.12, not 1809.12. The text in Section 5 states that SIRPF2 is 14.48 AIC units worse, which follows from the printed 1809.12 but not from the correct arithmetic, which gives a 10.48-unit difference. Since the AIC comparison is the central statistical evidence for the claim that SIRJPF2 is best, the table and all derived numerical comparisons need correction.
  2. [Supplement S7.1, Eq. (S16), and Table 2] The comparison against the Searle et al. (2016) model is not made under the same likelihood or fitting protocol: Supplement S7.1 describes scoring the Searle model using published predicted means and optimizing only the negative-binomial dispersion parameter tau. The statement in Section 5 that the Searle sexual-switch model is not a competitive statistical explanation therefore conflates model quality with fitting protocol. The authors should either refit the Searle model under the same observation model and likelihood as the other candidates, or explicitly restrict the claim 'better than any of the alternatives' to the models actually estimated under the same protocol (SIRPF2, SIRJPF2-Gamma, and the negative-binomial benchmarks).
  3. [Section 7 and Fig. 4] The paper correctly acknowledges in Section 7 that algal food levels were not measured and that the resource-depletion mechanism 'remains an unvalidated assumption in the theory rather than being directly supported by data.' Because this latent food dynamic is the basis for the paper's main biological interpretation in Section 5 and the simulated latent trajectories in Fig. 4, the contingency should be stated in the abstract and results as well as the discussion, and preferably supported by a sensitivity analysis under an alternative latent food specification. I regard this as a limitation rather than an internal inconsistency, but it is load-bearing for the biological conclusion.
minor comments (4)
  1. [Section 2] The text contains the typo 'treaments' in the description of the experimental design.
  2. [Section 5] The sentence introducing AIC comparisons contains the typo 'intepretable' and should read 'interpretable'.
  3. [Table 1] Several invasive-species parameters are reported with infinite or zero-bound confidence intervals (r_i, p_i, f_S^i); the text explains this identifiability issue, but a footnote in the table would help readers avoid misreading these as estimation failures.
  4. [Fig. 4] The caption's phrase 'dashed colored lines are simulations' is confusing for the algae and parasite panels, since those states have no observed data; clarify that the dashed lines are simulated latent trajectories from the fitted model.

Circularity Check

0 steps flagged · score 1.0 of 10

No circular derivation: the likelihood-based model comparison and out-of-sample juvenile validation are self-contained; the flagged Searle benchmark and latent-food limitations are correctness concerns, not circularity.

full rationale

The central statistical claim that SIRJPF2 fits better than the alternatives is supported by likelihood maximization on the adult Daphnia data, with juvenile data held out for genuine out-of-sample validation (Section 6; Fig. 4). The fitted model is compared against negative binomial polynomial benchmarks and against the nested SIRPF2 and SIRJPF2-Gamma variants under the same likelihood framework (Table 2), so the model-selection result does not reduce by construction to its inputs. The comparison to Searle et al. (2016) is scored from published predicted means with only the negative-binomial dispersion tau optimized (Supplement S7.1, Eq. S16); this is a legitimate benchmark-fairness concern that could affect the strength of the 'better than any alternative' claim, but it is not a circular reduction because those published means are external inputs, not parameters fitted and then renamed as predictions. The resource-depletion interpretation relies on the unmeasured latent food compartment F, and Section 7 explicitly concedes that this 'mechanism, although logical within the model's framework, remains an unvalidated assumption in the theory rather than being directly supported by data'; that weakens the biological conclusion but does not make the statistical fit tautological. Self-citations to Bretó et al. (2020) for panel iterated filtering, to Ionides et al. (2017) for MCAP, and to the authors' own software are methodology citations with published algorithms and code, and they are not the load-bearing evidence for the model comparison. Supplement S2 states that theoretical convergence results for MPIF 'are not yet available' and that support 'will be published elsewhere (Wheeler et al., 2025)', an in-preparation same-author work; this is an explicit missing-support flag, but the main AIC comparisons and diagnostics do not depend on that deferred theory, so it does not create circularity.

Assumptions & free parameters 10 free parameters · 6 assumptions · 2 invented entities

The model contains 26 fitted parameters plus several fixed biological constants, and the central resource-depletion interpretation rests on two unmeasured latent compartments. The scientific conclusion is therefore supported by statistical fit but not by independent measurement of the key mechanism.

free parameters (10)
  • r_n, r_i (birth rates) = 4.08e1, 2.15e5 (10^-6 cells^-1 L)
    Birth rates per unit food per adult; r_i is weakly identified with an upper CI of infinity.
  • fS_n, fS_i (susceptible adult filtration rates) = 1.10e-3, 2.42e-7 L individual^-1 day^-1
    Filtration rates; fS_i has CI (0, 9.47e-4), indicating weak identifiability.
  • p_n, p_i (infections per spore) = 2.72e-1, 1.34e3 (10^-3 individual spore^-1)
    Only products p*fS are tightly identified; individual p_i has huge uncertainty.
  • theta_S_n, theta_S_i, theta_I_n, theta_I_i, theta_J_n, theta_J_i, theta_P = see Table 1
    Mortality and removal rates; natural death rates are poorly identified in the parasitized treatment, while infected-host death rates are better identified.
  • xi (infected to susceptible filtration ratio) = 2.22e1
    Estimated above 1, contrary to individual-level expectation; reconciled by population-level genotypes and reverse causation.
  • sigma_I_n, sigma_I_i, sigma_J_n, sigma_J_i, sigma_F, sigma_P = 2.93e-4, 1.73e-7, 2.84e-1, 3.02e-1, 1.44e-1, 2.71e-1
    Process noise intensities; juvenile, food and parasite noise are important, while sigma_S is fixed at zero.
  • tau_S_n, tau_S_i, tau_I_n, tau_I_i = 4.10, 5.26, 9.02e-1, 1.39
    Negative-binomial measurement overdispersion parameters.
  • lambda_J, xi_J, beta_n, beta_i = 0.1, 1, 3e3, 3e3
    Fixed biological constants; lambda and xi_J are chosen by reasoning, beta is from prior parasite-yield literature.
  • mu, delta = 0.37, 0.013
    Algal refill and sampling rates fixed by the experimental protocol.
  • Initial latent state values F_0, P_0, J_0 = not reported in Table 1
    Initial state draws affect the likelihood; the protocol fixes some initial densities, but the paper does not state how initial values are estimated or randomized.
assumptions (6)
  • standard math Particle-filter estimates of the panel likelihood are approximately unbiased and converge as the number of particles grows.
    Invoked in Section 4 and Algorithm S1; standard sequential Monte Carlo theory, not developed in this paper.
  • domain assumption Mesocosm units are conditionally independent, so the panel likelihood factorizes as a product over units (Eq. 11).
    Experimental units are separate buckets; cross-unit contamination is not considered.
  • ad hoc to paper Algal food F and parasite spore P follow the specified latent SDEs (Eqs. 4 and 5) even though neither is measured.
    The food-depletion mechanism is load-bearing; no direct food measurements exist, so this is a model postulate.
  • domain assumption Juveniles mature at a fixed rate lambda = 0.1 per day and filter at the same rate as adults (xi_J = 1); juvenile data are excluded from fitting.
    Biologically motivated and sensitivity-checked in Section 6, but lambda_J is fixed, not estimated.
  • ad hoc to paper All parameters are shared across units, so replicate differences are explained by stochastic dynamics alone.
    Selected by AIC comparison with unit-specific alternatives in Table S1; if true heterogeneity exists, shared-parameter estimates could absorb it into noise.
  • domain assumption The Ito-interpreted Gaussian-noise SDE can be simulated by Euler-Maruyama without invalid nonnegativity problems.
    Section 3 and the gamma-noise variant in Section S9 suggest the Gaussian model is adequate, but the model does not structurally enforce nonnegativity.
invented entities (2)
  • Latent alga density F_u(t)
    purpose: Represents the food resource; depletion drives Daphnia decline under the proposed mechanism.
    Alga density was not measured in the analyzed data (Fig. 4 caption and Section 7), so its dynamics are inferred from the model rather than validated.
  • Latent parasite spore density P_u(t)
    purpose: Drives transmission; spore dynamics are unobserved but inferred from infection counts.
    The paper states that spores were not measured; the model output predicts a rise and plateau that is not empirically checked.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Mechanistic models for panel data: Analysis of ecological experiments with four interacting species." pith.science (2026). https://pith.science/paper/V3DE23LU

@misc{pith2026250604508,
  author       = {Pith},
  title        = {Pith review of: Mechanistic models for panel data: Analysis of ecological experiments with four interacting species},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/V3DE23LU}},
  note         = {Machine review of arXiv:2506.04508}
}
read the original abstract

In an ecological context, panel data arise when time series measurements are made on a collection of ecological processes. Each process may correspond to a spatial location for field data, or to an experimental ecosystem in a designed experiment. Statistical models for ecological panel data should capture the high levels of nonlinearity, stochasticity, and measurement uncertainty inherent in ecological systems. Furthermore, the system dynamics may depend on unobservable variables. This study applies iterated particle filtering techniques to explore new possibilities for likelihood-based statistical analysis of these complex systems. We analyze data from a mesocosm experiment in which two species of the freshwater planktonic crustacean genus, Daphnia, coexist with an alga and a fungal parasite. Time series data were collected on replicated mesocosms under six treatment conditions. Iterated filtering enables maximization of the likelihood for scientifically motivated nonlinear partially observed Markov process models, providing access to standard likelihood-based methods for parameter estimation, confidence intervals, hypothesis testing, model selection and diagnostics. This toolbox allows scientists to propose and evaluate scientifically motivated stochastic dynamic models for panel data, constrained only by the requirement to write code to simulate from the model and to specify a measurement distribution describing how the system state is observed.

Figures

Figures reproduced from arXiv: 2506.04508 by the authors.

Figure 1
Figure 1. The experimental data. Density (Individuals [PITH_FULL_IMAGE:figures/full_fig_p007_1.png] view at source ↗
Figure 2
Figure 2. A flow diagram for the SIRJPF2 model. The two species of [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. A flow diagram for panel iterated filtering. If the blue shaded blocks are omitted, [PITH_FULL_IMAGE:figures/full_fig_p016_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Simulated Daphnia densities (Individuals/Liter), alga density (106 ·cells/Liter) and parasite density (103 ·spores/Liter), over time (days), in the mixed species parasitized dynamics with parameters from [PITH_FULL_IMAGE:figures/full_fig_p025_4.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

53 extracted references · 51 canonical work pages

  1. [1]

    Akaike, H. (1974). A new look at the statistical model identification. IEEE Transactions on Automatic Control , 19(6):716--723

    1974

  2. [2]

    S., Maskell, S., Gordon, N., and Clapp, T

    Arulampalam, M. S., Maskell, S., Gordon, N., and Clapp, T. (2002). A t utorial on p article f ilters for o nline n onlinear/ n on- G aussian B ayesian t racking. IEEE Transactions on signal processing , 50(2):174--188

    work page 2002

  3. [3]

    A., Leos-Barajas, V., Mills Flemming, J., Nielsen, A., Petris, G., and Thomas, L

    Auger-M \'e th \'e , M., Newman, K., Cole, D., Empacher, F., Gryba, R., King, A. A., Leos-Barajas, V., Mills Flemming, J., Nielsen, A., Petris, G., and Thomas, L. (2021). A guide to state--space modeling of ecological time series. Ecological Monographs , 91(4):e01470

    work page 2021

  4. [4]

    L., Laneri, K., Pascual, M., Bouma, M., and Dhiman, R

    Bhadra, A., Ionides, E. L., Laneri, K., Pascual, M., Bouma, M., and Dhiman, R. C. (2011). Malaria in northwest I ndia: Data analysis via partially observed stochastic differential equation models driven by L \'e vy noise. Journal of the American Statistical Association , 106(494):440--451

    work page 2011

  5. [5]

    Bj rnstad, O. N. and Grenfell, B. T. (2001). Noisy clockwork: time series analysis of population fluctuations in animals. Science , 293(5530):638--643

    work page 2001

  6. [6]

    C., Streicker, D

    Blackwood, J. C., Streicker, D. G., Altizer, S., and Rohani, P. (2013). Resolving the roles of immunity, pathogenesis, and immigration for rabies persistence in vampire bats. Proceedings of the National Academy of Sciences , 110(23):20837–--20842

    work page 2013

  7. [7]

    Bret\' o , C. (2014). On idiosyncratic stochasticity of financial leverage effects. Statistics & Probability Letters , 91:20--26

    work page 2014

  8. [8]

    L., and King, A

    Bret \'o , C., He, D., Ionides, E. L., and King, A. A. (2009). Time series analysis via mechanistic models. The Annals of Applied Statistics , pages 319--348

    work page 2009

Show all 53 references
  1. [9]

    and Ionides, E

    Bret \'o , C. and Ionides, E. L. (2011). Compound M arkov counting processes and their applications to modeling infinitesimally over-dispersed systems. Stochastic Processes and their Applications , 121(11):2571--2591

    2011

  2. [10]

    L., and King, A

    Bret \'o , C., Ionides, E. L., and King, A. A. (2020). Panel data analysis via mechanistic models. Journal of the American Statistical Association , 115(531):1178--1188

    2020

  3. [11]

    A., and Ionides, E

    Breto, C., Wheeler, J., King, A. A., and Ionides, E. L. (2025). panel P omp: Analysis of panel data via partially observed M arkov processes in R . R Journal , to appear

    2025

  4. [12]

    T., Newman, K

    Buckland, S. T., Newman, K. B., Thomas, L., and Koesters, N. B. (2004). State-space models for the dynamics of wild animal populations. Ecological Modelling , 171(1-2):157--175

    2004

  5. [13]

    D., Van Damme, K., and Ebert, D

    Cornetti, L., Fields, P. D., Van Damme, K., and Ebert, D. (2019). A f ossil-calibrated p hylogenomic a nalysis of D aphnia and the D aphniidae. Molecular Phylogenetics and Evolution , 137:250--262

    2019

  6. [14]

    Coulson, T., Rohani, P., and Pascual, M. (2004). Skeletons, noise and population growth: The end of an old debate? Trends in Ecology and Evolution , 19:359--364

    2004

  7. [15]

    J., Anderson-Bergman, C., Lang, D

    de Valpine, P., Turek, D., Paciorek, C. J., Anderson-Bergman, C., Lang, D. T., and Bodik, R. (2017). Programming with m odels: w riting s tatistical a lgorithms for g eneral m odel s tructures with NIMBLE . Journal of Computational and Graphical Statistics , 26(2):403--413

    2017

  8. [16]

    Ebert, D. (2005). Ecology, Epidemiology, and Evolution of Parasitism in Daphnia . National Library of Medicine

    2005

  9. [17]

    J., Lachmann, M., Tec, M., Pasco, R., Woody, S., Du, Z., Wang, X., Ingle, T

    Fox, S. J., Lachmann, M., Tec, M., Pasco, R., Woody, S., Du, Z., Wang, X., Ingle, T. A., Javan, E., Dahan, M., Gaither, K., Escott, M. E., Adler, S. I., Johnston, S. C., Scott, J. G., and Meyers, L. A. (2022). Real-time pandemic surveillance using hospital admissions and mobil...

    2022

  10. [18]

    Havel, J. E. and Hebert, P. D. (1993). D aphnia lumholtzi in N orth A merica: Another exotic zooplankter. Limnology and Oceanography , 38(8):1823--1827

    1993

  11. [19]

    L., and King, A

    He, D., Ionides, E. L., and King, A. A. (2010). Plug-and-play inference for disease dynamics: Measles in large and small towns as a case study. Journal of the Royal Society Interface , 7:271--283

    2010

  12. [20]

    L., Bret \'o , C., and King, A

    Ionides, E. L., Bret \'o , C., and King, A. A. (2006). Inference for nonlinear dynamical systems. Proceedings of the National Academy of Sciences , 103(49):18438--18443

    2006

  13. [21]

    L., Breto, C., Park, J., Smith, R

    Ionides, E. L., Breto, C., Park, J., Smith, R. A., and King, A. A. (2017). Monte C arlo profile confidence intervals for dynamic systems. Journal of the Royal Society Interface , 14:1--10

    2017

  14. [22]

    L., Nguyen, D., Atchad \'e , Y., Stoev, S., and King, A

    Ionides, E. L., Nguyen, D., Atchad \'e , Y., Stoev, S., and King, A. A. (2015). Inference for dynamic and latent variable models via iterated, perturbed B ayes maps. Proceedings of the National Academy of Sciences , 112(3):719--724

    2015

  15. [23]

    Johnson, J. B. and Omland, K. S. (2004). Model selection in ecology and evolution. Trends in Ecology & Evolution , 19(2):101--108

    2004

  16. [24]

    A., Domenech de Cell \`e s, M., Magpantay, F

    King, A. A., Domenech de Cell \`e s, M., Magpantay, F. M. G., and Rohani, P. (2015). Avoidable errors in the modelling of outbreaks of emerging pathogens, with special reference to Ebola . Proceedings of the Royal Society of London, Series B , 282(1806):20150347

    2015

  17. [25]

    A., Nguyen, D., and Ionides, E

    King, A. A., Nguyen, D., and Ionides, E. L. (2016). Statistical inference for partially observed M arkov processes via the R package pomp. Journal of Statistical Software , 69(12)

    2016

  18. [26]

    Kitagawa, G. (1998). A self-organising state-space model. Journal of the American Statistical Association , 93:1203--1215

    1998

  19. [27]

    Kloeden, P. E. and Platen, E. (1999). Numerical Soluion of Stochastic Differential Equations . Springer, New York, 3rd edition

    1999

  20. [28]

    Kotov, A. A. and Taylor, D. J. (2011). Mesozoic fossils ( > 145 M ya) suggest the antiquity of the subgenera of Daphnia and their coevolution with chaoborid predators. BMC Evolutionary Biology , 11:1--9

    2011

  21. [29]

    W., Skaug, H., and Bell, B

    Kristensen, K., Nielsen, A., Berg, C. W., Skaug, H., and Bell, B. M. (2016). TMB : Automatic differentiation and L aplace approximation. Journal of Statistical Software , 70:1--21

    2016

  22. [30]

    L., King, A

    Li, J., Ionides, E. L., King, A. A., Pascual, M., and Ning, N. (2024). Inference on spatiotemporal dynamics for coupled biological populations. Journal of the Royal Society Interface , 21(216):20240217

    2024

  23. [31]

    A., Peacor, S

    Marino, J. A., Peacor, S. D., Bunnell, D. B., Vanderploeg, H. A., Pothoven, S. A., Elgin, A. K., Bence, J. R., Jiao, J., and Ionides, E. L. (2019). Evaluating consumptive and nonconsumptive predator effects on prey density using field times series data. Ecology , 100(3):e02583

    2019

  24. [32]

    A., and Rohani, P

    Martinez-Bakker, M., King, A. A., and Rohani, P. (2015). Unraveling the transmission ecology of polio. PLoS Biology , 13(6):e1002172

    2015

  25. [33]

    S., Clancey, E., McMichael, C., and Lofgren, E

    Mietchen, M. S., Clancey, E., McMichael, C., and Lofgren, E. T. (2024). Estimating sars-cov-2 transmission parameters between coinciding outbreaks in a university population and the surrounding community. medRxiv , pages 2024--01

    2024

  26. [34]

    J., Lavorel, S., Le Gall, L., Meslin, L., Morand, S., Morin, X., Morlon, H., Pinay, G., Pradel, R., Schurr, F

    Mouquet, N., Lagadeuc, Y., Devictor, V., Doyen, L., Duputi \'e , A., Eveillard, D., Faure, D., Garnier, E., Gimenez, O., Huneman, P., Jabot, F., Jarne, P., Joly, D., Julliard, R., K\' e fi, S., Kergoat, G. J., Lavorel, S., Le Gall, L., Meslin, L., Morand, S., Morin, X., Morlon...

    2015

  27. [35]

    J., Bilby, R

    Naiman, R. J., Bilby, R. E., Schindler, D. E., and Helfield, J. M. (2002). Pacific salmon, nutrients, and the dynamics of freshwater and riparian ecosystems. Ecosystems , 5(4):399--417

    2002

  28. [36]

    L., and Ritov, Y

    Ning, N., Ionides, E. L., and Ritov, Y. (2021). Scalable M onte C arlo inference and rescaled local asymptotic normality. Bernoulli , 27:2532--2555

    2021

  29. [37]

    ksendal, B. (1998). Stochastic Differential Equations . Springer, New York, 5th edition

    1998

  30. [38]

    D., Dorn, N

    Peacor, S. D., Dorn, N. J., Smith, J. A., Peckham, N. E., Cherry, M. J., Sheriff, M. J., and Kimbro, D. L. (2022). A skewed literature: Few studies evaluate the contribution of predation-risk effects to natural field patterns. Ecology Letters , 25(9):2048--2061

    2022

  31. [39]

    M., Lemanski, B

    Penczykowski, R. M., Lemanski, B. C., Sieg, R. D., Hall, S. R., Housley Ochs, J., Kubanek, J., and Duffy, M. A. (2014). Poor resource quality lowers transmission potential by changing foraging behaviour. Functional Ecology , 28(5):1245--1255

    2014

  32. [40]

    Radzikowski, J., Krupi \'n ska, K., and \'S lusarczyk, M. (2018). Different thermal stimuli initiate hatching of d aphnia diapausing eggs originating from lakes and temporary waters. Limnology , 19:81--88

    2018

  33. [41]

    Schnute, J. T. (1994). A general framework for developing sequential fisheries models. Canadian Journal of Fisheries and Aquatic Sciences , 51(8):1676--1688

    1994

  34. [42]

    L., Cortez, M

    Searle, C. L., Cortez, M. H., Hunsberger, K. K., Grippi, D. C., Oleksy, I. A., Shaw, C. L., de la Serna, S. B., Lash, C. L., Dhir, K. L., and Duffy, M. A. (2016a). Population density, not host competence, drives patterns of disease in an invaded community. The American Natural...

    2016

  35. [43]

    L., Shaw, C

    Searle, C. L., Shaw, C. L., Hunsberger, K. K., Prado, M., and Duffy, M. A. (2016b). Salinization decreases population densities of the freshwater crustacean, d aphnia dentifera . Hydrobiologia , 770:165--172

    2016

  36. [44]

    Simpson, M. J. and Maclaren, O. J. (2023). Profile-wise analysis: a profile likelihood-based workflow for identifiability analysis, estimation, and prediction with mechanistic mathematical models. PLoS Computational Biology , 19(9):e1011515

    2023

  37. [45]

    Subramanian, R., He, Q., and Pascual, M. (2021). Quantifying asymptomatic infection and transmission of COVID -19 in N ew Y ork C ity using observed cases, serology, and testing capacity. Proceedings of the National Academy of Sciences , 118(9)

    2021

  38. [46]

    Tan, K., Hooker, G., and Ionides, E. L. (2024). Accelerated inference for partially observed M arkov processes using automatic differentiation. arXiv:2407.03085

    2024 arXiv

  39. [47]

    Weiss, H. H. (2013). The SIR model and the foundations of public health. Materials Matematics , pages 1--17

    2013

  40. [48]

    Wen, L., Yin, Y., Li, Q., Peng, Z., and He, D. (2024). Modeling the co-circulation of influenza and COVID -19 in H ong K ong, C hina. Advances in Continuous and Discrete Models , 2024(1):32

    2024

  41. [49]

    J., and Ionides, E

    Wheeler, J., Abkemeier, A. J., and Ionides, E. L. (2025). Iterating marginalized B ayes maps for likelihood maximization. In preparation

    2025

  42. [50]

    Wheeler, J., Rosengart, A., Zhuxun, J., Tan, K. H. E., Treutle, N., and Ionides, E. L. (2024). Informing policy via dynamic models: Cholera in H aiti. PLOS Computational Biology , 20:e1012032

    2024

  43. [51]

    Wilkinson, D. J. (2018). Stochastic Modelling for Systems Biology . CRC press

    2018

  44. [52]

    Young, P. (1998). Data-based mechanistic modelling of environmental, ecological, economic and engineering systems. Environmental Modelling & Software , 13(2):105--122

    1998

  45. [53]

    L., Cortez, M

    Searle, C. L., Cortez, M. H., Hunsberger, K. K., Grippi, D. C., Oleksy, I. A., Shaw, C. L., de la Serna, S. B., Lash, C. L., Dhir, K. L., and Duffy, M. A. (2016). Population density, not host competence, drives patterns of disease in an invaded community. The American Naturali...

    2016

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.