The physical diversion had been stopped. It would have been convenient if the story had ended there, with a contractor accepting responsibility and a neighbor receiving an unpleasant letter from the district. But by then I had begun asking a question that bothered me more than the pipe itself.
How long had that water been missing before I heard it entering the lake?
Mara helped me answer it. She brought nine days of district turnout records and our meter readings into the same timeline. Every evening, shortly after my scheduled irrigation delivery began, the flow recorded at our orchard meter fell by roughly thirty-five to forty percent while the upstream meter remained within normal delivery range. Near sunrise the downstream flow rose again. There were slight variations according to which zones were open, but the pattern was too consistent to be an ordinary leak.
"They're filling at night," Sam said when we reviewed the graph.
"Someone was," Mara answered. "The telemetry proves where the loss was occurring in time, not who operated the valve."
She was right to be precise. The evidence did not yet identify the person who cut the lock or the person who ordered the nightly openings. It established that my paid irrigation water had been diverted for multiple turns, not only during the afternoon I discovered it.
I thought about the week before. I had blamed the slow moisture recovery in Block Five on the heat. We had adjusted zone schedules, cleaned filters, and spent extra hours making readings that still didn't add up. We had tried to compensate for a missing forty percent without knowing the water was leaving our system halfway uphill.
The realization made me angrier than Denise's words on the dock. She had not taken a bucket from a ditch. She had quietly inserted herself into the system and let us struggle with the consequences.
That night, Elena came to the farm store after her inspections. She had a folder of soil readings and photographs from the affected blocks. She told us she could not honestly estimate final crop loss while the fruit was still on the trees. The stress might show up in size, grade, and market value rather than in a dramatic row of dead trees.
"A tree can survive and still lose the fruit you planned to sell," she explained. "The mature Redhaven block may come out of this looking almost normal. Those three-year-old trees are another matter. We're going to track them carefully."
"What about the apples?" I asked.
"Honeycrisp is sensitive to stress during sizing. We need unaffected comparison rows and actual packout reports. Not photographs of a few small apples held in front of a camera."
I appreciated the bluntness. Elena had a degree in horticulture, years of field experience, and a complete lack of interest in giving me the answer I wanted just because I was paying her. She was the kind of expert I trusted, which meant she often told me things I did not enjoy hearing.