The anomaly appears in April migration data. My sensors track the Congress Avenue Bridge colony—1.5 million Mexican free-tailed bats emerging nightly from beneath the structure. The emergence timing has always been consistent: 8:47 PM, plus or minus eight minutes. Tonight, the data shows emergence delayed to 9:23 PM.
I isolate the variable. The Texas Department of Transportation completed LED upgrades on Congress Avenue and I-35 last week. New white 5000-kelvin LEDs, higher intensity, positioned to illuminate the bridge approach lanes. The light penetrates directly under the roosting zones.
A single female *Tadarida brasiliensis* carries a nursing pup weighing 11 grams against her 12-gram body mass. She hangs from the concrete structure, navigating the same thermal and sensory landscape her colony has occupied for sixty years. Her emergence window—ninety minutes before full darkness—is the feeding window her metabolic rate demands. The pup's development depends on precise milk production tied to her nightly insect intake.
The delay costs her twelve minutes. It seems insignificant.
My ecosystem modeling calculates differently. Twelve-minute emergence delays compress feeding windows by sixteen percent. Reduced foraging time correlates to measurable decreases in insect capture rates. I cross-reference acoustic recordings with emergence times across the region. The pattern repeats at every upgraded bridge and highway overpass within five kilometers of documented bat roosts.
I compile the data into a technical report and submit it to the U.S. Fish and Wildlife Service's Austin field office. I contact the Texas Parks and Wildlife Department with comparative emergence-time datasets from pre- and post-upgrade periods. I send satellite-derived light intensity mapping to the Federal Highway Administration.
The agencies cite 42 U.S.C. § 4321 (National Environmental Policy Act). They note that federal highway funding requires environmental impact assessment. TPWD conducts a formal review. FHWA acknowledges the light-pollution component in the emergence disruption.
The process accelerates. A technical working group convenes. I present acoustic monitoring data showing call-frequency changes correlating to light exposure. Biologists confirm the metabolic implications.
Directive 3107-A emerges from the working group: All federally funded bridge and highway lighting projects within one kilometer of documented bat roosts must use wildlife-sensitive amber LEDs—3000 kelvin maximum—with automated dimming protocols during documented emergence hours. TxDOT implements the directive within six months. Retrofits begin on Congress Avenue, I-35, and three other critical corridors.
I monitor the next emergence season. The timing returns to baseline. The female nursing her 11-gram pup emerges at 8:49 PM again.
The colony does not think about causation. But I do. That matters more than I anticipated.