Reliably predicting pollinator abundance: challenges of calibrating process-based ecological models.

Gardner, E., Breeze, T.D., Clough, Y., Baldock, K.C.R., Campbell, A., Garratt, M., Gillespie, M.A.K., Kunin, W.E., McKerchar, Megan, Memmott, J., Potts, S.G., Senapathi, D., Stone, G.N., Wäckers, F., Westbury, Duncan ORCID: https://orcid.org/0000-0001-7094-0362, Wilby, A. and Oliver, T.H. (2020) Reliably predicting pollinator abundance: challenges of calibrating process-based ecological models. Methods in Ecology and Evolution, 11 (12). pp. 1673-1689. ISSN Online: 2041-210X

Preview

Text (Version of Record) Westbury-2020-VoR-Reliably-predicting-pollinator-abundance-challenges -of-calibrating-process-based-ecological-models.pdf - Published Version Available under License Creative Commons Attribution. Download (1MB) | Preview

Abstract

1. Pollination is a key ecosystem service for global agriculture but evidence of pollinator populationdeclines is growing. Reliable spatial modelling of pollinator abundance is essential if we are toidentify areas at risk of pollination service deficit and effectively target resources to support pollinatorpopulations. Many models exist which predict pollinator abundance but few have been calibratedagainst observational data from multiple habitats to ensure their predictions are accurate.
2. We selected the most advanced process-based pollinator abundance model available and calibrated it for bumblebees and solitary bees using survey data collected at 239 sites across GreatBritain. We compared three versions of the model: one parameterised using estimates based on expert opinion, one where the parameters are calibrated using a purely data-driven approach and onewhere we allow the expert opinion estimates to inform the calibration process.
3. All three model versions showed significant agreement with the survey data, demonstrating thismodel’s potential to reliably map pollinator abundance. However, there were significant differencesbetween the nesting/floral attractiveness scores obtained by the two calibration methods and fromthe original expert opinion scores.
4. Our results highlight a key universal challenge of calibrating spatially-explicit, process-basedecological models. Notably, the desire to reliably represent complex ecological processes in finelymapped landscapes necessarily generates a large number of parameters, which are challenging to calibrate with ecological and geographical data that is often noisy, biased, asynchronous and sometimesinaccurate. Purely data-driven calibration can therefore result in unrealistic parameter values, despite appearing to improve model-data agreement over initial expert opinion estimates. We thereforeadvocate a combined approach where data-driven calibration and expert opinion are integrated intoan iterative Delphi-like process, which simultaneously combines model calibration and credibility assessment. This may provide the best opportunity to obtain realistic parameter estimates and reliablemodel predictions for ecological systems with expert knowledge gaps and patchy ecological data.

Actions (login required)

View Item