First Fieldwork: Installing EchoQuilt Without Disturbing Clusters
The One Visit That Decides the Whole Winter
An early September morning in 2023, a five-person bat-crew in Ulster County approached the west entrance of a 14,000-bat Myotis lucifugus and Indiana bat hibernaculum to deploy sensors before swarm intensified. The team had rehearsed the sequence on a dry run the week before: decontaminate boots and bags, enter in two pairs spaced 12 minutes apart, unpack sensors with gloves on, place them by pre-marked rock features rather than by GPS, leave, and document everything with a single wide-angle photograph taken from the entrance. Total in-cave time per person: 38 minutes. The sensors then ran unattended for 218 days, including the entire critical disturbance window of Nov 1-Apr 1. That single install was the only human presence until April retrieval.
This is the disturbance tradeoff that makes the whole EchoQuilt approach work: one careful entry before hibernation earns you a full winter of passive data. Get it wrong and you either disturb a swarming colony or find in January that your array is silent. Installation entries can trigger arousals with measurable mass loss even in pre-hibernation when bats are still active, so the protocol needs to respect that the disturbance budget starts the moment boots enter the cave.
Pre-swarm timing is not arbitrary. By early September, most northeastern hibernacula have a small but rising population of Myotis and Perimyotis arriving from summer maternity sites. By late September, swarming intensifies as mating activity ramps up. By early October, torpor begins for the earliest-arriving individuals. The install window threads between these phases: late enough that the hibernaculum's airflow patterns have stabilized into their winter regime, but early enough that no torpid clusters have formed and the few bats present are mobile and capable of flying away from disturbance. Miss the window early and the airflow regime is still shifting; miss it late and you walk past forming clusters.
State DNR bat-crew coordinators should set their install dates against site-specific historical swarm and torpor data, not against a regional default — the difference between the right and wrong install week can be the difference between an arousal-free winter and a compromised dataset.
Pre-Swarm Install, Patch by Patch
EchoQuilt's install workflow is designed around one rule: the first-year quilt must be stitched from a single entry. The sensor kit is built to match. Each recorder weighs under 200 g, runs on lithium cells rated for 240+ days at 2-10°C, and mounts to a friction clamp that grips speleothem or bolted anchor without adhesives or percussion. A standard four-sensor configuration covers a 25 m chamber; an eight-sensor configuration covers a multi-room hibernaculum up to 80 m long. The recorders are identical except for their position, which means a biologist does not need to match sensor-to-location by role.
The installation sequence has five patches that stitch into the quilt's first frame. First, entrance airflow node: placed inside the throat, at least 1 m from the drip line, oriented along the prevailing airflow axis. Second, geometry anchors: two recorders at 2-3 m above floor level, on opposite walls, in the chamber where the largest expected cluster forms. Third, a cluster-proximate recorder at least 3 m from the nearest expected roost surface, never directly under. Fourth, a deep-chamber recorder in the quietest back volume for microclimate baseline. Fifth, optional low-frequency geophones on bedrock for micro-tremor sources. Each patch is logged in the quilt as an independent source of acoustic illumination, and the install photograph becomes a permanent registration frame.
Decontamination follows the national WNS protocol exactly: hot water at or above 55°C for 20 minutes on every piece of gear, including sensors, before entering. A parallel set of decontaminated retrieval gear is prepped now and stored separately so that April retrieval uses virgin-clean equipment. The USFWS Range-wide Indiana bat and NLEB guidelines govern permit timing and the allowed entry windows, so biologists should have their Section 7 paperwork reviewed by their USFWS field office before scheduling. EchoQuilt sensor housings are designed to tolerate the 55°C decon bath without degrading microphone calibration; the housings drain in 30 seconds and re-calibrate at room temperature within an hour, which fits into a pre-entry staging window without slowing the team.
The actual deployment borrows from established passive acoustic workflow. Wildlife Acoustics' deployment guide for acoustic bat surveys covers mounting height, calibration pings, and environmental noise verification; EchoQuilt uses the same discipline with two differences. The mounts hold tighter tolerances because the geometry inversion is sensitive to sensor drift, and a 2 mm shift in any sensor across a winter measurably degrades cluster-boundary precision. And the calibration is done outside the cave with a reference generator before entry, so the in-cave time holds to the minutes-not-hours budget. This first-visit discipline is the practical expression of the low-disturbance basics philosophy underlying the entire monitoring program.
Advanced Tactics for Clean First-Year Deployments
Three tactics separate a clean first-year deployment from a compromised one. First, run a pre-install acoustic dry run on the outside of the cave. Set the full sensor array under a shelter 100 m from the entrance for 48 hours and verify every channel is recording properly, clocks are synchronized, and the file-writing cadence matches the plan. Half of all first-year data gaps come from unit-level failures that a dry run catches before entry; catching them in the field means a second entry, which is exactly what the workflow is trying to avoid. The dry run also exposes battery anomalies — a sensor that fails a 48-hour outdoor run will not survive 218 days underground at 4°C, where lithium chemistry slows further. Replace any borderline unit before it crosses the cave threshold.
The dry run also seeds the disturbance-free logging workflow that governs ongoing monitoring, since the same sensor mounts and acoustic baselines used for first-year geometry will support guano pile tracking, microclimate fusion, and species occupancy across subsequent winters.
Second, stage the install in two non-overlapping entry windows separated by at least a week. Pre-place mounts and clamps during a mid-August visit when bats are minimally present, then return in early September with calibrated recorders. Splitting the high-manipulation work (drilling, clamping) from the low-manipulation work (mounting an already-calibrated unit) means each entry is shorter and less invasive. The mid-August window also lets the team scout for new breakdown, water levels, or entrance changes that might force a sensor relocation before September — better to reroute on paper than under a swarming colony.
Third, document everything with the acoustic quilt itself. The first hour of recording after sensors are powered up becomes the baseline geometry frame, and team movement inside the cave during install is logged by the array as a disturbance signature that the quilt can subtract from subsequent data. This gives biologists a clean start for the long winter, informed by the same first-responder onboarding discipline used in cross-niche rapid deployments. A useful side effect is that the install-disturbance signature itself becomes a reference dataset for what a small human party sounds like in the chamber, which improves later detection of unauthorized entry events during the unattended winter window.
Get Early Access to EchoQuilt
If you are planning pre-swarm hibernaculum deployments for the 2026-27 season and need a low-disturbance install workflow that your state DNR bat crew or USFWS field office will sign off on, EchoQuilt is now actively accepting applications for the first-year pilot cohort. We provide the sensor kit, a fully site-specific install plan reviewed in advance against your Section 7 paperwork, and direct support from our field team during the September install window. The pilot also includes a post-install acoustic review at 30 days and 90 days so any drift in sensor performance is caught while it can still be corrected from the surface, not after April retrieval shows the data was incomplete. Join the Waitlist for Hibernacula Biologists to start scoping sensor counts, permit timing, and decontamination logistics for your site.