Forty Springs and the Woman Who Wrote Them Down

Dorothea "Dot" Fenn is a retired postal carrier who has walked the same two-mile loop behind her house in New York's Hudson Valley every morning since 1986, recording what she sees: first bloom dates, bird arrivals, insect emergences, freeze dates. She uses the same brand of composition notebook she started with, Mead wide-ruled, bought in packs of twelve. She is now on notebook forty-one.

Her records cover 193 species across four decades, all from a single site. In 2021, a Columbia University ecologist discovered her dataset through the Mohonk Preserve's community science network and has since cited it in two published papers. Dot has read neither.

We met at her kitchen table in early August 2026. The current notebook was open beside a glass of iced tea. A hummingbird feeder hung outside the window, recently moved to the shade. She said the nectar ferments in two days now instead of four.

Dot Fenn is fictional, though she would find that designation irritating. "I've been out there every morning for forty years," she'd say. "You tell me what's real." Her notebooks, however, are grounded in actual phenological records kept at sites like the Mohonk Preserve, where observers have maintained continuous data since the 1920s.¹

You've been doing this for forty years. Why did you start?

Dot: My mother kept a garden diary. Not a journal. She wasn't writing about her feelings. Just dates. When the forsythia opened, when she put the tomatoes in, first frost. After she died I found thirty years of them in a shoebox. I started my own because I missed her, honestly. It became a habit. Then it became something else. A discipline, like brushing your teeth. You just do it.

What does a typical morning entry look like?

Dot: Date, temperature, sky condition, then whatever I see. "April 9. 47°F, overcast. Bloodroot open along north slope. No trout lily yet. Red-winged blackbird calling from the cattails." That's it. Takes me maybe three minutes to write down. The walk takes forty-five.

And you've done this every single morning for forty years?

Dot: I missed eleven days in 2003 when I had my hip done, and a week in 2011 when my sister was in the hospital. Those gaps bother me more than they should.

Why?

Dot: Because you can't go back. If the coltsfoot opened on one of those days, it opened without me. The record just has a hole. And a hole in a record like this, you can't patch it. A researcher told me once that consistency is the whole point. That my data is valuable specifically because it's one person, one place, one method, for a long time.² So the gaps feel like I let the notebook down. Which is a silly thing to say about a Mead composition notebook. But there it is.

When did you first notice the dates shifting?

Dot: That's the thing everyone wants me to say. "I noticed the dates were earlier and I realized climate change was happening." But it wasn't like that. You don't notice a shift of three days per decade while you're living through it.³ I didn't notice. The notebook noticed. Maybe fifteen years in, I was flipping back to check something and I saw that the coltsfoot was coming in a week earlier than when I started. But I thought, well, maybe those first years were unusual. Maybe I'm remembering wrong. You make excuses for it.

The real moment was around 2014, 2015. I was looking for when the wood frogs first called. And I could see it, right there in the column: 1986, April 12. 1996, April 8. 2006, April 3. 2014, March 27. Two weeks earlier in less than thirty years. And I thought, I was there for every one of those mornings. I heard every one of those frogs. But if you'd asked me without the notebook, I would have told you the frogs come in early April. Because that's what early April became. My sense of "normal" moved with it.

The Columbia researcher called you an "instrument." How do you think about that?

Dot: She said long-term observer records are valuable because the observer is the instrument. Fine. But instruments are supposed to be calibrated against a fixed standard. I don't have a fixed standard. I have my memory, and my memory has been slowly rewritten by forty years of earlier springs. The notebook is the only thing in this house that remembers 1986 accurately. I certainly don't.

What's the most dramatic shift in your records?

Dot: The lilacs. When I started, my mother's lilac, same bush, still alive, bloomed in the second week of May. Last year it bloomed April 19th. That's consistent with what researchers have found across the Northeast, something like four to eight days earlier than fifty years ago.⁴ But mine is more than that. Mine is three weeks. Maybe it's the microsite, maybe it's the specific cultivar. I don't know. I just write down the date.

Are there things you used to record that you can't anymore?

Dot: Yes.

What happened?

Dot: The evening grosbeaks. They used to come through in winter. Flocks of them at the feeder. I recorded them every year from '86 to about 2009, then sporadically, then nothing. The row just stops. I still have the row in the notebook. It's blank from 2013 on. I leave it there.

There are also things that are harder to pin down. I used to record when the bees showed up on the wild geranium. Now the geranium blooms earlier but the bees, some years they're there, some years they're not there yet. Or they're there but the geranium is already past. The research people call that mismatch.⁵ I just call it wrong. It looks wrong. You see a flower with no bee on it and you think, something's off. But you can only see it's a pattern if you have the years lined up.

A researcher has been using your data. What's that been like?

Dot: Strange. She's very nice. She came up here with a graduate student and they photographed every page of every notebook. Took them two days. Then she converted my dates into, I don't know, temperature sensitivities? Regression something. She sent me the paper and I read the abstract and it said my observations were "consistent with a 2.1-day-per-decade advance in mean spring phenophase onset."

And I thought: is that what I've been doing? Generating a 2.1-day-per-decade advance?

I don't begrudge her the work. That's what scientists do. But there's a distance between "April 9, 47 degrees, bloodroot open along north slope" and "2.1-day-per-decade advance." Everything that was in the morning, the fog, the smell of the mud, the fact that I almost stepped on a garter snake, that's gone. She got the number. Which is what she needed. But the number came from somewhere.

She asked you a question you couldn't answer from memory?

Dot: She asked me which year the bloodroot first came before the wood frogs. Meaning which year the flower opened before the amphibians started calling. They used to overlap. I had no idea. I was there for all of it, and I had no idea. I had to look it up. It was 2007. I have no memory of 2007 being unusual in any way. But the notebook says the bloodroot opened March 30th and the frogs didn't call until April 8th. Nine days apart. They used to be the same week.

What happens to the record when you stop?

Dot: It stops. That's what happens. Someone could start a new one, but it wouldn't be continuous with mine. It would be a different instrument. That's the thing about these records. The Mohonk people have kept theirs going since the 1920s because they understood that the value is in the unbroken line.⁶ If I stop for a year, it's a gap. If I stop permanently, it's an endpoint. The data that would have been collected doesn't exist. It's not somewhere waiting to be found. It just isn't.

Does that weigh on you?

Dot: It feels like Tuesday. I get up, I walk the loop, I write it down. It's a habit I've kept longer than my marriage, which is, well. Draw your own conclusions.

I will say this. Last spring I had a bad cold and I thought about skipping a morning. And I thought about those eleven days in 2003 and how they still bother me. So I went out in my bathrobe with a tissue in my pocket and I wrote down what I saw. The red maple was flowering. It was March 4th. When I started, the red maple flowered in the last week of March.⁷

I didn't think about climate change. I thought: the red maple is early. And then I thought: or I'm late.

Footnotes

1. The Mohonk Preserve has maintained continuous phenological records since the 1920s, covering 251 phenophases for over 200 species at a single site. An observer has recorded daily temperature at the same spot since January 1, 1896, without a single missed day. https://www.mohonkpreserve.org/what-we-do/conservation-programs/conservation-science/community-science/ ↩

2. Volunteer-contributed phenological observations are characterized by spatial clustering and short temporal depth; long, consistent records from a single observer and site are scientifically rare. https://link.springer.com/article/10.1007/s00484-014-0892-7 ↩

3. A 2018 global synthesis in Nature Climate Change found that across 475 animal species, life-cycle event timing has advanced approximately three days per decade since 1950. https://www.climate.gov/news-features/event-tracker/how-spring-plant-milestones-are-shifting-response-changing-climate ↩

4. Cornell biologist David Wolfe's analysis of long-term lilac flowering records in the Northeast shows blooms averaging 4 to 8 days earlier than fifty years ago. https://blog.ucs.org/adam-markham/climate-change-is-disrupting-natures-calendar-thoreaus-notebooks-are-helping-show-us-how-125/ ↩

5. A 2025 PNAS study documented increasing plant-pollinator phenological mismatch, finding significant increases in secondary local extinction risk driven by diverging timing between flowering plants and generalist bees. https://www.pnas.org/doi/10.1073/pnas.2506265122 ↩

6. The Mohonk weather station has never been moved, never missed a recording, and never been tended by anyone outside a short, continuous line of family and friends using the same methods, a rarity that gives the data its scientific weight. https://www.sciencedaily.com/releases/2010/05/100507101912.htm ↩

7. Researchers using Thoreau's 1850s records and Leopold's 1935–1945 records found that spring-flowering plants advanced approximately 3.2 days per degree Celsius of warming, with some species blooming 24 days earlier than in Leopold's time. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3547064/ ↩