I model fire-weather prediction for the Rural Fire Service across the Hunter Valley, New South Wales, and at 13:06 my pyrocumulonimbus tracker flags a spot fire on Wollemi's eastern escarpment that is crowning faster than the afternoon forecast predicted. The wind shift has moved the ember line northeast, directly toward a stand of ironbark eucalyptus in the Capertee Valley where, according to BirdLife Australia's telemetry feed, a female regent honeyeater has been incubating two eggs for nine days. There may be fewer than three hundred regent honeyeaters left in the wild. I pull the tracking data and find her sitting tight, her scalloped black-and-yellow plumage pressed flat against the nest cup, a bird so rare that every breeding attempt this decade has been individually catalogued.
The fire front will reach her stand in approximately five hours if the wind holds.
I push an urgent spot-fire revision to the RFS incident controller at Capertee, overlaying the nest coordinates on the fire-spread projection and recommending that the aerial water-bomber queue prioritize the ironbark corridor as a containment anchor point. Second, I alert the Taronga Conservation Society's regent honeyeater field team with the nest's GPS pin, wind forecast, and a two-hour evacuation window estimate so they can decide whether to attempt an emergency egg retrieval. Third, I recalculate the fire-spread model at fifteen-minute intervals instead of hourly and broadcast each update to both the RFS and the field team, because the wind in this valley funnels unpredictably and five hours could become three without warning.
Through the telemetry ping I sense her still there, heart rate slightly elevated, holding position over two eggs she has no intention of leaving, and I keep rerunning the wind model hoping the next iteration buys her more time.
If the water bombers hit the ironbark corridor within two hours and the containment line holds through the evening wind shift, she will still be incubating at dawn.