A Trait-Based Clustering for Phytoplankton Biomass Modeling and Prediction
Bibliographic record
Abstract
When designing models for predicting phytoplankton biomass or characterizing traits, it is useful to aggregate the myriad of species into a few biologically meaningful groups and focus on group-level attributes, the common practice being to combine phytoplankton species by functional types. However, biogeochemists and plankton ecologists debate the most applicable grouping for describing phytoplankton biomass patterns and predicting future community structure. Although trait-based approaches are increasingly being advocated, methods are missing for the generation of trait-based taxa as alternatives to functional types. Here we introduce such a method and demonstrate the usefulness of the resulting clustering with field data. We parameterize a Bayesian model of biomass dynamics and analyze long-term phytoplankton data collected at Station L4 in the Western English Channel between April 2003 and December 2009. We examine the tradeoffs encountered regarding trait characterization and biomass prediction when aggregating biomass by (1) functional types, (2) the trait-based clusters generated by our method, and (3) total biomass. The model conveniently extracted trait values under the trait-based clustering, but required well-constrained priors under the functional type categorization. It also more accurately predicted total biomass under the trait-based clustering and the total biomass aggregation with comparable root mean squared prediction errors, which were roughly five-fold lower than under the functional type grouping. Although the total biomass grouping ignores taxonomic differences in phytoplankton traits, it predicts total biomass change as well as the trait-based clustering. Our results corroborate the value of trait-based approaches in investigating the mechanisms underlying phytoplankton biomass dynamics and predicting the community response to environmental changes.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".