That afternoon Maria walked the exposed stretch alone. She pictured the older man Ruth had described sitting on a hill through a storm, looking for ways to guide water without fighting its weight. His work had survived in fragments, yet the county no longer remembered why it existed. Maria thought about how quickly knowledge could vanish when it remained attached to one person instead of to a written plan. She decided that whatever else happened to the ranch, the next person would not have to begin with a mysterious line of greener grass.
She bought waterproof survey flags and a second notebook. The first remained her daily work record, filled with expenses, weather, broken equipment, and the small disasters of maintaining the place. The second was devoted to the drainage system. Each page had a location, measurements, a sketch, and a question. When a question was answered, she dated the answer rather than erasing the earlier uncertainty, because she wanted to be able to see how her understanding changed.
The road department confirmed that the existing outlet needed to remain clear and that proposed changes could not send concentrated water onto the county roadway. The conservation office helped her identify a modest restoration sequence that avoided major excavation and emphasized existing flow paths. None of this made the project cheaper. It did prevent her from turning a neglected system into a new problem in the name of rescuing it.
Maria began with a short stone-lined stretch above the worst waterlogged area. She removed accumulated silt by hand, preserving the surviving stones wherever she could and placing the sediment well away from the channel so the next rain would not carry it straight back. Some stones shifted when she touched them, exposing gaps where roots had grown through the original bed. She reset loose pieces with care, using the slope readings to preserve a gradual flow instead of creating a sudden drop. The work progressed so slowly that some evenings she seemed to have changed only a few feet of ground.
In one section she discovered that the stones were not laid across the bottom of a continuous open ditch, as she had first imagined. They formed low edges and occasional covered passages beneath places that had once been used for crossing. Beyond them the water had apparently been allowed to spread through shallow vegetated runs rather than remain confined to a single narrow trench. This explained why the green strip widened and narrowed: the system had different roles at different points. It also explained why a careless modern plow could have damaged it without leaving an obvious scar.
She made mistakes. At the first small outlet she removed too much accumulated soil, and a trial flow from a water tank moved faster than expected. Instead of soaking gradually through the prepared section, it cut a tiny groove at the edge of the channel and began carrying loose particles toward the next low spot. Maria shut the valve, filled the groove with stable material, and sat on the ground feeling foolish. She had spent weeks warning herself not to hurry, then created a problem by opening a section only a few inches deeper than necessary.