Pooled survival permitted both signs of an annual ozone timing link in a 1992-1993 Canadian extract

We tested whether a public ozone extract supports a link between development time and survival in the same taxon and exposure year. It has 22 timing rows from six papers and 36 survival or mortality rows from ten papers after dispersal removal. Only two papers have both classes. One pairs caterpillar timing with parasitoid survival. The other gives annual timing but two-year mean survival. No exact group has annual data for both outcomes. Under equal year weights, exact pooled means and annual survival between zero and 100 percent, the larval survival response in 1992 minus that in 1993 ranges from -176 to +176 percentage points. Both signs of its annual association with timing are possible. Quarter and three-quarter weights give the same sign uncertainty. These are sharp possibility bounds, not confidence intervals. They do not establish actual survival values or absence of a biological link. Missing annual denominators, selected reporting and one Canadian paper limit inference. Regional and period replication are not supported. Code, aggregate evidence, source hashes, feasible examples and independent arithmetic checks are included. Climate and weather interactions were not tested.

ozone · development time · survival · ecological timing · observation bias · partial identification · temporal aggregation