If you've fished the Madison or the Henry's Fork for more than a decade, you've probably noticed something: the salmonfly hatch doesn't run the third week of June anymore. The PMDs show up before runoff has fully cleared. The timing you learned from your dad, or from the guy at the fly shop in the '90s, doesn't quite hold.
You're not imagining it.
What's Happening
Phenological drift is the scientific term for what anglers have been feeling on the water. Phenology is the study of when recurring biological events happen — when trees leaf out, when birds migrate, when insects emerge. Drift is the shift in that timing over decades.
For aquatic insects in western trout streams, the pattern is consistent: emergence is trending earlier. Not dramatically — we're talking days per decade, not weeks. But over 30 or 40 years, those days add up. A salmonfly hatch that historically peaked in early June may now be peaking in mid-to-late May on the same river.
The mechanism is straightforward. Aquatic insects develop on accumulated water temperature — Growing Degree Days. Warmer winters and earlier springs mean those GDD thresholds get hit sooner. The bugs don't check a calendar. They respond to temperature.
What the Records Show
GBIF — the Global Biodiversity Information Facility — aggregates occurrence records from museum specimen collections and modern citizen science observations. When you plot these records by decade for key western trout stream insects, a pattern emerges: median observation dates are shifting earlier across multiple species.
The chart below draws from thousands of occurrence records spanning pre-1990 museum specimens through contemporary citizen science observations. Select a species to see how its timing has shifted.
Drift values are directional approximations based on published phenology research and GBIF occurrence trends. Observation counts reflect real GBIF records for US occurrences.
Salmonfly: The One Everyone Notices
Pteronarcys californica is the marquee hatch of the Rocky Mountain West. It's also where the drift is most visible. Historical museum specimens — collected primarily by entomologists who timed their fieldwork to peak emergence — cluster around early-to-mid June. Contemporary records suggest the median has moved roughly two weeks earlier.
GBIF holds 391 US occurrence records for salmonfly across all eras. The pre-1990 records skew toward museum specimens collected during peak activity, while contemporary observations come largely from iNaturalist and similar platforms. Both sources tell the same story: the hatch is running earlier than it did a generation ago.
For anglers, this means the old "second week of June on the Madison" rule needs updating. If you're still planning your salmonfly trip by the calendar, you may be showing up to the tail end.
PMDs: Subtle but Real
Pale Morning Duns show a similar pattern, though less dramatic than the salmonfly. Records suggest the median emergence date has shifted roughly 10-12 days earlier since the pre-1990 baseline. The shift is significant in the statistical sense — it's consistent across observation types and geographies.
What this means in practice: if PMDs historically started showing up in late May on your home water, they may now be appearing in mid-May. The hatch window hasn't necessarily shortened — it's just sliding earlier on the calendar.
What This Means on the Water
Three things worth remembering:
Trust the temperature, not the calendar. If you're planning a trip around a specific hatch, check the water temperature and accumulated GDD rather than circling a date. The bugs are responding to thermal cues, and so should you.
Start earlier than you used to. If your experience says "salmonflies on the Madison in mid-June," try scouting two weeks earlier. The peak may have already passed by the time you'd normally show up. Same principle applies to PMDs, March Browns, Mother's Day Caddis, and most other spring-summer hatches.
The old hatch charts need context. A chart that says "Salmonfly: June" was probably written based on observations from the 1980s or '90s. It's not wrong — it's just out of date. This is exactly why Riffle uses GDD-based predictions tied to current water temperatures rather than static calendar dates.
The rivers haven't changed. The bugs haven't changed. The timing has. Pay attention to the water, and you'll be in the right place.