The Outer Atoll was four miles north of the main Lagoon. Koa covered the distance in eleven minutes, pacing himself at a steady cruising speed, listening as he traveled. The acoustic character of the water changed as he moved north.
Inside the Lagoon and on the main reef, the reef structure produced a rich, layered sound environment — many surfaces, many depths, many sizes of organism contributing to a complex acoustic picture. As he moved into the open water between the main reef and the atoll, the sound simplified. Less structure, more open water, the reef behind him now a source of reflected signals rather than a surrounding environment.
Ahead of him, the atoll began to resolve from its general low-frequency wash into something with more detail. He could hear the individual impact signatures of wave energy striking the atoll's outer face. He could hear the difference between sections where the rock was solid and sections where it was porous, the porous sections producing a higher-frequency buzzing undertone as water forced through their tiny internal channels.
He was listening for something different from either of these. He was listening for a rhythmic pulse — the signature of water moving through a gap large enough to pass a dolphin-sized body, small enough to channel the flow rather than simply letting it spread. He reached the outer edge of the atoll's underwater base and turned east, following the base of the structure.
At depth, the atoll was a wall of rock and ancient coral that descended from the surface forty feet above him to the reef floor twenty feet below. He clicked steadily as he moved, sending sonar pulses into the rock face and reading the returns. Solid rock: a sharp, clean return signal.
Porous rock: a diffused return with scattered secondary signals. Open water: no return, or a delayed return from far surfaces. He moved east along the base for three minutes.
Mostly solid rock, some porous sections, no significant gaps. He turned and went west from his starting point, moving carefully, clicking every few seconds. Six minutes west.
Nothing. He surfaced briefly to breathe, then dove again and moved deeper, ten feet below his previous search depth, following the atoll base at the thirty-foot level. At four minutes west of center, he found it.
The return signal at that position was completely different from everything on either side. Where solid rock returned a clean echo and porous rock returned a scattered one, this section returned something that his sonar had no framework for — a delayed, channeled return that came back from two separate distances simultaneously, as though the signal had traveled through the wall and reflected off something on the far side before returning to him.
He hovered in front of the gap and clicked at maximum frequency. The return confirmed it. A passage.
He could not see it yet — the entrance was below him, oriented slightly downward into the base of the atoll structure rather than horizontal. He descended and found the mouth. It was narrower than he had expected from the acoustic signature.
About five feet wide at the entrance, roughly oval, the bottom covered in fine sand that had been moved into the gap by years of water flow. He hovered at the entrance and sent a long series of clicks through it. The passage went straight for approximately eight body-lengths, then turned — he could tell because the return signal shifted direction partway through — then straightened again.
The far end opened into what his sonar read as open water. The Blue Haven Lagoon. He clicked for the passage dimensions as accurately as he could.
Five feet wide at the entrance. The narrowest section, near the turn, was slightly less — four and a half feet. An adult dolphin in normal relaxed posture was approximately three feet across at the widest point of the body.
The adult males in the pod were larger. But every adult female and every calf could pass through four and a half feet of width if they moved carefully. He needed to know the surge cycle.
He held position at the entrance and felt the water. The pulse Marina had described was there — a rhythmic pressure increase as swell energy moved through the passage, then a release, then another increase. He timed it.
Fourteen seconds between pressure peaks. The peak itself lasted about four seconds, during which the water in the passage moved at a speed that would require active effort to swim against. Ten seconds of manageable flow between peaks.
Enough for a swimmer moving at speed. Enough, barely, for a calf swimming alongside its mother. He mapped everything he could measure about the passage with his sonar, storing the data in the spatial memory that dolphins maintained for important locations the way humans stored a map image.
Then he turned and headed back toward the Lagoon to report. Behind him, the water to the south had taken on a faint metallic taste that had not been there an hour before. The bloom was moving faster than Marina had estimated.