Updated Mar 22, 2026 • ~6 min read
Chapter 10: Losing Ground
FINN
She had a secondary camera.
He found it on the third day — Thursday morning, when he was checking the north path before he came to the cottage for the boat access session. He’d stopped at the post to confirm the primary camera’s position — it was back on the upper headland angle, which meant she’d reset it after he’d left on Wednesday — and then he’d seen the second unit at the base of the adjacent post, partially tucked behind the coastal shrub.
It was aimed at the north path. The field of view, based on the lens angle, would cover the stretch from the lighthouse junction to the post itself.
He stood on the path and looked at the secondary camera.
He thought: *she set this after I moved the first one.*
He thought: *she had it in her equipment from the beginning.*
He thought: *she was planning for exactly this.*
He reached for the camera.
His eagle stopped him.
Not physically. His eagle couldn’t stop him physically. But there was a quality to the attention — the focused, amused quality that it had been directing at this researcher since her truck came up the coast road — that said, in whatever register his eagle communicated in: *if you move this one, she’ll know.*
He thought: *she already knows.*
He thought: *she set up the secondary because she knows about the first.*
He thought: *if I move this one, she will have footage of me moving it.*
His eagle was amused in a way that he found deeply unhelpful.
He stepped back from the camera.
He drove down to the cottage.
She was at the dock when he arrived — she’d walked the south path from the cottage, the route Cal had told her about, and come down the alternate access to the dock rather than waiting for him at the cottage. She had her full field pack, camera, binoculars, a tide table printed from the morning’s data, and the specific expression of someone who had planned this morning in detail.
He said: “You found the south path.”
She said: “Cal told me about it. It shaves fifteen minutes off the cottage-to-dock route.” She looked at the boat. “Are we going out?”
He said: “Conditions are good.”
They went out.
The boat was the community’s work vessel — a fifteen-foot fiberglass hull with a covered bow for gear storage, reliable in the coastal swell conditions that October produced. He ran it north along the cliff base to the outer rock access point, the route that gave the best approach to the outer face’s lower nest sites, while she documented the cliff from the water with the camera.
She was good on the boat. Not performatively comfortable — genuinely comfortable, the kind of settled body posture that came from time spent on working water rather than recreational boats. She braced correctly when the swell hit the bow. She didn’t grab for the sides.
He said: “You’ve been on boats.”
She said: “Fieldwork on coastal rivers and lakes for seven years. I prefer canoes but I’ll take what the survey requires.”
He said: “Canoes.”
She said: “Quieter for close approach. On this kind of coast you’d want something larger.” She was photographing the cliff face as they moved. “There — the north-facing site behind the rock column. I couldn’t see that from any of the cliff path positions.”
He looked where she was pointing.
The north-facing site had four individuals. She was already writing.
He ran the boat along the outer base while she counted the outer face in sections, building her grid systematically. She called out individuals and positions, using the rock features as reference points, and wrote in the notebook in a hand that was clearly practiced at boat-condition field notes — small, anchored letters that didn’t depend on a stable surface.
She got to forty-four.
She looked up.
She said: “Forty-four. From the outer face plus the east and south bay counts there’s significant overlap — I’ve got confirmed individual-level identifications for twenty-three and I’ve seen individuals from the cliff path in recognisable positions here.”
He said: “The population range extends across the full headland.”
She said: “Yes.” She was looking at the cliff face with the expression of someone working through a problem. “The accessible positions gave me thirty-nine distinct observations. The outer face gives me additional individuals. The database figure is forty-one.”
He said: “Some of those forty-four will be the same individuals you counted from the accessible positions.”
She said: “Some of them. Twenty-three confirmed identifications I’ve carried across. That leaves twenty-one unidentified from the outer face count.” She looked at her notes. “I’ve got six confirmed new individuals — markings I haven’t recorded before. That would put my full-survey count at forty-five, possibly forty-six.”
He said: “Above the database figure.”
She said: “Possibly. I need to run the deduplication analysis on the identification data.” She looked at him. “The database says forty-one.”
He said: “The database has been counting from the accessible positions.”
She said: “Yes.” She looked at the outer face. “The outer face adds individuals the accessible count misses. The database figure of forty-one may be an undercount.”
He said: “It may be.”
She said: “So the population may be larger than the database shows.”
He said: “Possible.”
She wrote something in the notebook.
He watched her write and thought: *she’s building toward something and the outer face count just changed the direction.*
The database showed forty-one. If she came back with forty-five or forty-six, the question shifted from *why is the mortality rate zero* to *why does the database undercount the population*. Which was a different question. A less dangerous question, in some ways.
He thought: *she’s not going to stop at the count.*
She said: “I want to run the full deduplication analysis tonight. Can we go back tomorrow for the north section? I missed the very north end — there’s an overhang I couldn’t see from the boat at this tide.”
He said: “Tomorrow’s conditions are good. Same time.”
She said: “Thank you.”
They headed back to the dock.
She said, without looking up from her notebook: “I also want to access the upper headland section.”
He said: “The upper headland section is above the cliff path network. There’s no maintained access.”
She said: “I saw a path from the north cliff position. It looked maintained.”
He said: “Unmaintained track.”
She said: “I’d like to use it.”
He said: “I’ll need to check conditions.”
She said: “When.”
He said: “Early next week.”
She wrote something.
He thought: *she’s writing: early next week.*
His eagle found this funny.
He was losing this at a rate that was, at this point, statistically significant. He had moved a camera. She had set a secondary. He had redirected three separate questions about the upper headland. She had noted each redirect.
He thought: *I need to talk to Maren.*
He thought: *I need to decide what I’m actually doing.*
He thought: *my eagle decided three days ago.*
He thought: *that’s the problem.*
Reader Reactions