Insights into the quantification and reporting of model-related uncertainty across different disciplines

Simmonds, Emily G.; Peprah, Adjei, Kwaku; Andersen, Christoffer Wold; Aspheim, Janne Cathrin Hetle; Battistin, Claudia; Bulso, Nicola; Christensen, Hilary; Cretois, Benjamin; Cubero, Ryan John; Davidovich, Iván A.; Dickel, Lisa; Dunn, Benjamin; Dunn‐Sigouin, Etienne; Dyrstad, Karin; Einum, Sigurd; Giglio, Donata; Gjerløw, Haakon; Godefroidt, Amélie; González-Gil, Ricardo; Cogno, Soledad Gonzalo; Große, Fabian; Halloran, Paul R.; Jensen, Mari F.; Kennedy, John; Langsæther, Peter Egge; Laverick, Jack H.; Lederberger, Debora; Li, Camille; Mandeville, Elizabeth G.; Mandeville, Caitlin P.; Moe, Espen; Schröder, Tobias; Nunan, David; Parada, Jorge Sicacha; Simpson, Melanie Rae; Skarstein, Emma Sofie; Spensberger, Clemens; Stevens, Richard; Subramanian, Aneesh; Svendsen, Lea; Theisen, Ole Magnus; Connor, Watret,; O’Hara, Robert B.

doi:10.1016/j.isci.2022.105512

Cited by 11 publications

(15 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For predicting global vegetation patterns it is of fundamental importance to know whether the ecosystem state at a particular location in environmental space can be predicted based on external forcing by climate system variables or whether positive feedbacks acting on the initial ecosystem state might determine ecosystem state ( 19 , 20 ). Here we revisit the challenge posed by the alternative ecosystem state hypothesis for prediction models, by using a recently proposed method for characterizing the influence of climate on vegetation ( 22 ).…”

mentioning

confidence: 99%

Limited climatic space for alternative ecosystem states in Africa

Higgins

Conradi

Kruger

et al. 2023

Science

View full text Add to dashboard Cite

One of the foundational premises of ecology is that climate determines ecosystems. This has been challenged by alternative ecosystem state models, which illustrate that internal ecosystem dynamics acting on the initial ecosystem state can overwhelm the influence of climate, and by observations suggesting that climate cannot reliably discriminate forest and savanna ecosystem types. Using a novel phytoclimatic transform, which estimates the ability of climate to support different types of plants, we show that climatic suitability for evergreen trees and C4 grasses are sufficient to discriminate between forest and savanna in Africa. Our findings reassert the dominant influence of climate on ecosystems and suggest that the role of feedbacks causing alternative ecosystem states is less prevalent than has been suggested.

show abstract

mentioning

confidence: 99%

Limited climatic space for alternative ecosystem states in Africa

Higgins

Conradi

Kruger

et al. 2023

Science

View full text Add to dashboard Cite

show abstract

“…Uncertainty is a key element of the scientific process but one that can often be overlooked (Simmonds et al., 2022), with potentially detrimental consequences (Aeberhard et al., 2018; Edeling et al., 2021; Regan et al., 2005). In this study, we explored the contemporary standards of uncertainty reporting and propagation in papers using MPMs, as well as the consequences of omission.…”

Section: Discussionmentioning

confidence: 99%

“…The high prevalence of reporting at least some uncertainty in MPM papers (78%) is a commendable achievement. This subfield performs substantially better at reporting parameter‐related uncertainty than other areas in ecology (Simmonds et al., 2022). Nevertheless, this reporting prevalence is still lower than in other fields such as political or health science (Simmonds et al., 2022) and, therefore, there is room for improvement, as a fifth of MPM papers report no demographic rate uncertainty at all.…”

Section: Discussionmentioning

confidence: 99%

“…Such differential methodological practices and different data availability could lead to missing uncertainty estimates for particular stages, ages or demographic rates; for example, uncertainty may be reported for a mark-recapture survival analysis but not for mean reproductive success (Paez et al, 2015). Omissions of key model-related uncertainties are common in ecology and evolution (Simmonds et al, 2022), and we might expect the same within MPM studies.…”

Section: Introductionmentioning

confidence: 99%

“…When insights from such models are used to inform action, as exemplified by the use of predictive models during the COVID‐19 pandemic (Edeling et al., 2021; James et al., 2021; Kreps & Kriner, 2020; Meehan et al., 2020) or in fisheries management (Aeberhard et al., 2018; Harwood & Stokes, 2003; Kindsvater et al., 2018), the consequences of an incorrect or incomplete consideration of uncertainty include economically costly sub‐optimal action (Aeberhard et al., 2018), damaged credibility (Kreps & Kriner, 2020; Palliser & Dodson, 2017) or even loss of life (Edeling et al., 2021; Regan et al., 2005). Despite these potentially severe consequences, model‐related uncertainties are not yet consistently or completely considered in ecological studies (Simmonds et al., 2022).…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Uncertainty propagation in matrix population models: Gaps, importance and guidelines

Simmonds,

Jones

2023

Methods Ecol Evol

Self Cite

View full text Add to dashboard Cite

Matrix population models (MPMs), which describe the demographic behaviour of a population based on age or stage through discrete time, are popular in ecology, evolutionary biology and conservation biology. MPMs provide a tool for guiding management decisions and can give insight into life history trade‐offs, patterns of senescence, transient dynamics and population trajectories. These models are parameterised with estimates of demographic rates (e.g. survival and reproduction) and can have multiple layers of underlying statistical analyses, all of which introduce uncertainty. For accurate and transparent results, this uncertainty should be propagated through to quantities derived from the MPMs, such as population growth rates (λ). However, full propagation is not always achieved, leading to omitted uncertainty and negative consequences for the reliability of inferences drawn. We summarised the contemporary standards regarding demographic rate uncertainty reporting and propagation, by reviewing papers using MPMs from 2010 to 2019. We then used reported uncertainties as the basis for a simulation study to explore the impact of uncertainty omission on inferences drawn from the analysis of MPMs. We simulated four scenarios of demographic rate propagation and evaluated their impact on population growth rate estimates. Although around 78% of MPM papers report some kind of uncertainty in their findings, only half of those report uncertainty in all aspects. Additionally, only 31% of papers fully propagate uncertainty through to derived quantities. Our simulations demonstrate that, even with moderate levels of uncertainty, incomplete propagation introduces bias. Omitting uncertainty may substantially alter conclusions, particularly for results showing small changes in population size. Biased conclusions were most common when uncertainty in the most influential demographic rates for population growth were omitted. We suggest comprehensive guidelines for reporting and propagating uncertainty in MPMs. Standardising methods and reporting will increase the reliability of MPMs and enhance the comparability of different models. These guidelines will improve the accuracy, transparency and reliability of population projections, increasing our confidence in results that can inform conservation efforts, ultimately contributing to biodiversity preservation.

show abstract