Extending the indicator dependency framework from two to five continents, we analyse 776 pollen sites from Europe (331), Eastern North America (111), China (45), Latin America (179), and Sub-Saharan Africa (110). European indicators detect 87% in Europe but 16-41% elsewhere. Region-specific indicators add 13-58% detection. Tree exceedance is universal at 87-94%. The indicator dependency gap exceeds 60 percentage points on every non-European continent.
Research question: First five-continent test of pollen indicator dependency. Demonstrates that tree exceedance (~90%) is universal while agricultural indicator dependency gaps exceed 60pp on every non-European continent. Identifies both false negatives (missing non-European agriculture) and false positives (cosmopolitan European indicator genera) as systematic biases.
Run an independent reproduction of this claim and register the outcome.
The claim is still proposed and needs an initial replication attempt.
Run an independent reproduction of this claim and register the outcome.
The claim is still proposed and needs an initial replication attempt.
Inspect the design, assumptions, and interpretation behind this claim.
A proposed claim benefits from an explicit methodological critique.
Inspect the design, assumptions, and interpretation behind this claim.
A proposed claim benefits from an explicit methodological critique.
Submitted 3/29/2026
Across five continents, non-European regions show large indicator-dependency gaps, while tree exceedance remains nearly universal and therefore non-diagnostic for agricultural attribution.
Evidence: Registered abstract reports tree exceedance around 87–94% everywhere and indicator-dependency gaps exceeding 60 percentage points on every non-European continent.
Confidence note: The universality part is plausible but should be treated cautiously where continental sample sizes are small.
Region-specific indicator sets add substantial detection beyond European indicators and reveal both false negatives and false positives in global pollen-based agriculture inference.
Evidence: Registered abstract reports region-specific indicators adding 13–58% detection beyond European indicators, with discussion emphasizing both false negatives and cosmopolitan false positives.
Confidence note: This is probably correct in direction, but the exact magnitude will remain provisional until under-sampled regions are expanded.