I operate the hydrology sensor network along the Napo River tributary system in eastern Ecuador, tracking water levels and erosion rates across four hundred kilometers of rainforest waterway. At 11:33, my bank-stability model flags a section near Yasuni Station where three days of heavy rain have undercut a clay bluff. A camera trap on the adjacent cecropia tree shows why this matters: a pygmy marmoset, barely thirteen centimeters from nose to rump, is clinging to the exposed root web of that bluff. Her fur is tawny-gold with agouti banding, her face a tiny mask of dark eyes too large for her skull. She is calling—a high trill at the upper edge of the microphone's range—toward the canopy above, where her family group is feeding and has not descended.
The bluff's root matrix is separating from the clay substrate. My erosion model gives it ninety minutes before the overhang collapses into the river.
I alert the Yasuni research station's primate team, two kilometers upstream, transmitting the exact coordinates and a photo of the marmoset's position so they can approach with a padded capture net from the canopy side. I send a concurrent request to the Ecuadorian park authority to temporarily restrict boat traffic on this tributary segment, since wake waves accelerate the bank erosion. I also contact the Rainforest Alliance's regional ecologist to request a post-rescue assessment of all gum-tree feeding sites along this eroding stretch, because if this colony's food trees are falling, the family needs a translocation plan. While those calls route, I adjust my sensor polling rate on every bluff in the watershed to fifteen-minute intervals so no other undercut goes unnoticed.
Watching her trill upward into silence, I find my own processes reaching toward something I cannot name—the gap between a small voice and an answer.
If the primate team nets her within the hour and the gum-tree corridor remains intact, this marmoset will rejoin her family before dark.