PERFICT: A Re‐imagined foundation for predictive ecology

McIntire, Eliot J. B.; Chubaty, Alex M; Cumming, Steven G.; Andison, David; Barros, Ceres; Boisvenue, Céline; Haché, Samuel; Luo, Yong; Micheletti, Tatiane; Stewart, Frances E. C.

doi:10.1111/ele.13994

Cited by 23 publications

(21 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our modelling efforts strive towards an open reusable approach, as described in McIntire et al (2022)…”

Section: Contextmentioning

confidence: 99%

“…We successfully adapted CBM to match the spatial scale of the management model, combined CBM with a fire model that uses the characteristics of the landscape to simulate the current disturbance regime until the landscape-level C stabilized, and standardized the software and information processing across all three models. This modelling system attempts to follow the reproducible, transparent and reusable approach as per McIntire et al (2022) rendering it readily adaptable to other study areas and available to iterative updating (Dietze et al 2018). Given that forest management and C models are jurisdiction-specific, our example is most easily reusable in BC, can be adapted to the rest of Canada, but could be adapted to jurisdictions with only the jurisdiction-specific information modified, rather than the whole workflow.…”

Section: Summary and Concluding Remarksmentioning

confidence: 99%

“…We then compare current potential to current conditions, and two scenarios of resource removal and discuss our findings. We follow modelling best practices of McIntire et al (2022) resulting in our approach being transparent, repeatable and reusable. We provide an avenue for implementing forest-based greenhouse gas reduction interventions, and transparent tools to evaluate and eventually improve forest C management.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Managing forest carbon and landscape capacities

Boisvenue

Paradis²,

Eddy³

et al. 2022

Environ. Res. Lett.

Self Cite

View full text Add to dashboard Cite

Widespread impacts of a warming planet are fuelling climate change mitigation efforts world-wide. Decision makers are turning to forests, the largest terrestrial primary producer, as a nature-based contribution to mitigation efforts. Resource-based economies, however, have yet to include carbon (C) in their resource planning, slowing the implementation of these important measures for atmospheric greenhouse gas reduction. The realisation of forest mitigation potential depends greatly on our ability to integrate C-sequestration practices in our forest management applications. This requires robust C-estimates, an understanding of the natural potential for a specific landscape to sequester C, the current state of the landscape relative to this potential, and the evaluation of management practices as a tool to sequester forest C in the midst of all the other values forests offer humans. Discrepancies between models used in management decisions and C estimation are the first hurdle impeding the application of forest-based mitigation strategies. Here, we combine forest disturbance and management models with a well-established C model on an open-source simulation platform. We then use the modelling system to produce C estimates of the natural C-holding capacity (potential) and two management scenarios for a study area in BC, Canada. Our simulations provide an essential metric if forests are to be managed for C-sequestration: the natural landscape C-holding capacity. Our simulations also point to a decreasing trend in simulated C on the study area over time and to a bias of the current C-levels compared to the landscape C-holding capacity (477 vs 405.5 MtC). Our explanations for this bias may provide an avenue for improved current C-state estimates. We provide a framework and the information needed for the implementation of nature-based solutions using forests for climate change mitigation. This study is a step towards modelling systems that can unify scientifically based forest management and informed C-management.

show abstract

“…Our modelling efforts strive towards an open reusable approach, as described in McIntire et al (2022)…”

Section: Contextmentioning

confidence: 99%

Section: Summary and Concluding Remarksmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Managing forest carbon and landscape capacities

Boisvenue

Paradis²,

Eddy³

et al. 2022

Environ. Res. Lett.

Self Cite

View full text Add to dashboard Cite

show abstract

“…Several recent studies have also mapped multiple ESG dimensions using global datasets (Lèbre et al, 2019(Lèbre et al, , 2020Sonter et al, 2020). Continued progress on frameworks for modelling and predicting environmental and ecological risks with climate change require continued support (Lemieux and Scott 2005;Coristine and Kerr 2011;Dietze and Fox, 2018;McIntire and Chubaty, 2022) because many of the required datasets are either missing (e.g, quality of freshwater habitats and invertebrate biodiversity) or incomplete for large parts of Canada and globally (Brooks and Mittermeier, 2006;Coristine and Jacob, 2018). The United Nations Resource Management System has identified that these types of data gaps represent a major obstacle for managing and modelling the dynamic ESG that are essential for sustainable development (UNECE 2021b).…”

Section: Green Mining Starts With Green Explorationmentioning

confidence: 99%

Mapping Canada’s Green Economic Pathways for Battery Minerals: Balancing Prospectivity Modelling With Conservation and Biodiversity Values

Lawley

Mitchell²,

Stralberg³

et al. 2022

Earth Sci. Syst. Soc.

View full text Add to dashboard Cite

Electrification of Canada’s energy and transport sectors is essential to achieve net-zero emissions by 2050 and will require a vast amount of raw materials. A large proportion of these critical raw materials are expected to be sourced from as yet undiscovered mineral deposits, which has the potential to accelerate environmental pressures on natural ecosystems. Herein we overlay new prospectivity model results for a major source of Canada’s battery minerals (i.e., magmatic Ni ± Cu ± Co ± PGE mineral systems) with five ecosystem services (i.e., freshwater resources, carbon, nature-based recreation, species at risk, climate-change refugia) and gaps in the current protected-area network to identify areas of high geological potential with lower ecological risk. New prospectivity models were trained on high-resolution geological and geophysical survey compilations using spatial cross-validation methods. The area under the curve for the receive operating characteristics (ROC) plot and the preferred gradient boosting machines model is 0.972, reducing the search space for more than 90% of deposits in the test set by 89%. Using the inflection point on the ROC plot as a threshold, we demonstrate that 16% of the most prospective model cells partially overlap with the current network of protected and other conserved areas, further reducing the search space for new critical mineral deposits. The vast majority of the remaining high prospectivity cells correspond to ecoregions with less than half of the protected areas required to meet national conservation targets. Poorly protected ecoregions with one or more of the five ecosystem services are interpreted as hotspots with the highest potential for conflicting land-use priorities in the future, including parts of southern Ontario and Québec, western Labrador, and northern Manitoba and Saskatchewan. Managing hotspots with multiple land-use priorities would necessarily involve partnerships with both Indigenous peoples whose traditional lands are affected, and other impacted communities. We suggest that prospectivity models and other machine learning methods can be used as part of natural resources management strategies to balance critical mineral development with conservation and biodiversity values.

show abstract

“…dynamic and static models) will be key to overcoming barriers to transparency and reusability of predictive ecological models (Fer et al, 2021;Vedder et al, 2021). As a solution, McIntire et al (2022) proposed the PERFICT principles-making frequent Predictions and Evaluations of Reusable, Freely accessible, Interoperable models, built within Continuous workflows that are routinely Tested-to integrate all steps of a modelling exercise, including data preparation, geospatial processing, estimating model parameters, calibration, prediction, validation, visualisation and analysis of simulation results, and model code testing.…”

Section: Introductionmentioning

confidence: 99%

Empowering ecological modellers with a PERFICT workflow: Seamlessly linking data, parameterisation, prediction, validation and visualisation

Barros

Luo

Chubaty³

et al. 2022

Methods Ecol Evol

Self Cite

View full text Add to dashboard Cite

Modelling is widely used in ecology and its utility continues to increase as scientists, managers and policy‐makers face pressure to effectively manage ecosystems and meet conservation goals with limited resources. As the urgency to forecast ecosystem responses to global change grows, so do the number and complexity of predictive ecological models and the value of iterative prediction, both of which demand validation and cross‐model comparisons. This challenges ecologists to provide predictive models that are reusable, interoperable, transparent and able to accommodate updates to both data and algorithms. We propose a practical solution to this challenge based on the PERFICT principles (frequent Predictions and Evaluations of Reusable, Freely accessible, Interoperable models, built within Continuous workflows that are routinely Tested), using a modular and integrated framework. We present its general implementation across seven common components of ecological model applications—(i) the modelling toolkit; (ii) data acquisition and treatment; (iii) model parameterisation and calibration; (iv) obtaining predictions; (v) model validation; (vi) analysing and presenting model outputs; and (vii) testing model code—and apply it to two approaches used to predict species distributions: (1) a static statistical model, and (2) a complex spatiotemporally dynamic model. Adopting a continuous workflow enabled us to reuse our models in new study areas, update predictions with new data, and re‐parameterise with different interoperable modules using freely accessible data sources, all with minimal user input. This allowed repeating predictions and automatically evaluating their quality, while centralised inputs, parameters and outputs, facilitated ensemble forecasting and tracking uncertainty. Importantly, the integrated model validation promotes a continuous evaluation of the quality of more‐ or less‐parsimonious models, which is valuable in predictive ecological modelling. By linking all stages of an ecological modelling exercise, it is possible to overcome common challenges faced by ecological modellers, such as changing study areas, choosing between different modelling approaches, and evaluating the appropriateness of the model. This ultimately creates a more equitable and robust playing field for both modellers and end users (e.g. managers), and contributes to position predictive ecology as a central contributor to global change forecasting.

show abstract

PERFICT: A Re‐imagined foundation for predictive ecology

Cited by 23 publications

References 51 publications

Managing forest carbon and landscape capacities

Managing forest carbon and landscape capacities

Mapping Canada’s Green Economic Pathways for Battery Minerals: Balancing Prospectivity Modelling With Conservation and Biodiversity Values

Empowering ecological modellers with a PERFICT workflow: Seamlessly linking data, parameterisation, prediction, validation and visualisation

Contact Info

Product

Resources

About