Towards improving species distribution models and vulnerability assessments of Canadian birds
Bibliographic record
Abstract
Climate change is set to impact biodiversity around the globe. In response, pole-wards range shifts are being observed ubiquitously, leading to range contraction for species already living at northern latitudes, such as birds breeding in northern Canada. Any pole-wards shift by these species, or from other species moving up from the south, will cause range contraction and increased extinction risk. Canadian birds are experiencing climate-driven change faster than anywhere else in the world, making it critical to understand the magnitude of change they will experience in the future. Species distribution models (SDMs) are the most commonly used tool to understand the incoming climate-driven changes. Inferring from associations between observational occurrences and environmental data, these models are used to predict potential species distribution with future climate scenarios. However, many different distribution models have been developed, each based on different assumptions and better suited for different types of data. The Canadian north is a particular challenge, as it has rapid climate change and very sparse distribution data. While most SDMs are based only on occurrence data, some new approaches combine occurrences with abundances from systematic survey data, which could be a solution to having reliable models in under-sampled regions. In addition to projected species range shifts, species traits are well-known to be correlates for extinction risk. Combining SDMs with traits could provide a framework for understanding a species ability to cope or adapt to climate-driven change as well as changes in habitat suitability and identify vulnerable species not currently deemed at-risk.In this thesis, I first address the question of how to integrate climate-change projections into a trait vulnerability assessment (TVA) framework. This new framework evaluates how much climate change a species is experiencing (i.e., how much suitable habitat they are predicted to lose and gain), which is then combined with species-specific traits. Species traits represent sensitivity, exposure, and adaptive capacity to climate change. By incorporating both SDM and TVA, I assessed the overall vulnerability of the 471 birds breeding in Canada and highlighted 83 species not currently at-risk, but likely to become vulnerable in the future given their combined changing distributions and capacity to withstand these changes.Secondly, I ask how different data types can be leveraged to address data-deficiencies when predicting species distributions. I test a recently developed method of combining abundance and occurrence data for waterfowl of the western boreal region of Canada, where both types of data are limited and biased in different ways. I compare four different types of data and approaches including: (1) abundance data from the Waterfowl Breeding Population and Habitat Survey (WBPHS), (2) occurrence data derived from abundance data from WBPHS, (3) occurrence data weighted by abundance, and (4) occurrence data from the Global Biodiversity Information Facility (GBIF). I find that the simple method of model integration (occurrence data weighted by abundances) produces better predictions than individual models. I also determine which models are most appropriate depending on species rarity.Overall, I find that an improved understanding of extinction risk is possible even in the rapidly-changing under-studied Canadian north, but we must leverage all available information to have reliable predictions of species risk
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 machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.009 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.002 | 0.001 |
| Scholarly communication | 0.003 | 0.002 |
| Open science | 0.003 | 0.002 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.002 | 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 source (direct Gemma or distilled Codex), 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".