Setting minimum head capsule abundance and taxa deletion criteria in chironomid-based inference models |
| |
| Authors: | Roberto Quinlan John P Smol |
| |
| Institution: | (1) Paleoecological and Environmental Assessment Research Laboratory (PEARL), Department of Biology, Queen's University, Kingston, Ontario, Canada, K7L 3N6 |
| |
| Abstract: | Criteria for removing training set lakes and taxa in chironomid based inference models, due to low abundances, have largely been ad hoc. We used an anoxia inference model and a hypolimnetic oxygen model from southcentral Ontario to determine what effect subfossil head capsule abundance and taxa deletion criteria have on fossil inference statistics. Results from six training set lakes suggest that a minimum abundance of 40–50 head capsules is sufficient for use in inference models, however more diverse samples likely require more than 50 head capsules. Taxa deletion criteria substantially improved the predictive ability of inference models (lowered the root mean squared error of prediction (RMSEP)). The common practice of including taxa with only  2% abundance in at least two lakes was one of the deletion criteria that much improved inference models. Similar deletion criteria, such as 2% in at least 3 lakes and 3% in at least 1 lake, produced comparable improvements (up to 18% reduction in RMSEP). |
| |
| Keywords: | Chironomidae inference models minimum counts head capsules taxa deletion methods counting procedures |
| 本文献已被 SpringerLink 等数据库收录！ |
|
