A small number of wildlife studies have run continuously for decades, and they produce findings no short project can reach. The scarcity of such studies is a matter of institutional structure rather than scientific interest.
Why duration changes what can be asked
Populations fluctuate for reasons that have nothing to do with the trend a researcher is looking for, including weather, disease and random variation in births and deaths.
A three-year study cannot distinguish a real decline from an ordinary trough, because both look identical over that span. Only a longer series establishes what normal variation looks like.
Long studies also capture rare events. Droughts, disease outbreaks and unusually severe winters reveal how a population responds to stress, and they cannot be scheduled.
Funding runs on short cycles
Research grants are typically awarded for a few years, and each renewal is competitive and judged partly on novelty rather than continuity.
Continuing an established dataset is precisely the kind of proposal that reads as unambitious, so long studies are often sustained by patching together small grants from many sources.
A single gap in funding can end a series permanently, because the value lies in the continuity and a break cannot be filled retrospectively.
Career incentives point the other way
Early-career researchers need publications within a few years, which favours projects designed to produce results quickly rather than to contribute a further year to someone else's record.
Long studies also distribute credit awkwardly. The person who started a dataset may not be present for the findings it eventually supports, and the person maintaining it may not have designed it.
Institutions rarely have a mechanism to inherit a study when its founder retires, so many end at that point despite the data remaining valuable.
Consistency is harder than it sounds
Comparability across decades requires that methods stay unchanged, even when better methods become available, because a change in technique can create an apparent trend.
Studies that adopt new equipment therefore run both methods in parallel for a period to calibrate the difference, which costs effort that produces no new finding.
Observers change too, and different people count and estimate differently. Training protocols and repeated cross-checks exist to keep that drift within known bounds.
What the surviving studies have shown
Long records are where shifts in breeding timing, gradual range movements and the delayed effects of habitat change have actually been demonstrated rather than inferred.
They also produce individual life histories, following animals from birth to death and revealing how early conditions shape reproduction decades later.
These findings are the reason such studies are defended so strongly by ecologists, and the reason their loss is not recoverable by starting a new one.