Effects of Anthropogenic Noise from Oil and Gas Development on Bat Activity in the Piceance Basin

Trent Hawkins & Dr. Rachel Buxton Colorado State University: Department of Fish, Wildlife, and Conservation Biology

Methods ●  18 ultrasonic recorders (model SM2, Wildlife Acoustics) were deployed in

the Piceance Basin, north-west Colorado, from May to September 2015. The number of nights recorded at each site ranged from 14-155 days.

●  Recording sites were far from oil and gas infrastructure (>1 km – control) or at varied distances <1 km from active (noise treatment) or dormant wells

Objectives This study addresses three questions:

1) Does oil and gas infrastructure affect the distribution of bats in the Piceance Basin? 2) Does bat activity increase as time since well drilling ceased increases? 3) Does noise from the remaining infrastructure

affect bat activity?

Pallid Bats (Antrozous pallidus) (left) and the Mexican Free-tailed bats (Tadarida brasiliensis) (right) are two species commonly found in the Piceance Basin.

Background ●  North American bat populations face an array of

threats, including: white nose syndrome, habitat loss, and climate change1.

●  Constant noise produced by active oil and gas wells can interfere with the echolocation calls of bats with low-frequency echolocation calls (<35 kHz), leading to decreased foraging rates and therefore, decreased activity2,3.

●  The Piceance Basin supports one of the largest

natural-gas reserves in North America and thus has a high density of energy infratructure4.

●  16 bat species have been recorded in the Piceance, with high potential overlap between bat habitat and oil and gas wells5.

●  Oil and gas development and activity in the

Piceance Basin has steadily declined since 2010 due to reduced global oil prices. This presents a unique opportunity to study both the recovery of bats, as well as the residual effects that energy infrstructure might have on bat activity.

Future Directions ●  Investigate how oil and gas infrastructure affects species

specific bat activity. ●  Examine how the amount of time since drilling noise ceased

affects the relationship between distance to well and activity. ●  Examine if remaining noise affects hourly or nightly bat activity.

References and Acknowledgements 1Warner, K. A. (2016.). Investigating The Effects of Noise Pollution From Energy Development on the Bat Community in The Piceance Basin. Retrieved February 3, 2017. 2Bunkley, J. P., & Barber, J. R. (2015). Noise Reduces Foraging Efficiency in Pallid Bats ( Antrozous pallidus). Ethology, 121(11), 1116-1121. 3Bunkley, J. P., Mcclure, C. J., Kleist, N. J., Francis, C. D., & Barber, J. R. (2015). Anthropogenic noise alters bat activity levels and echolocation calls. Global Ecology and Conservation, 3, 62-71. 4Lendrum, P. E., Anderson Jr, C. R., Monteith, K. L., Jenks, J. A., & Bowyer, R. T. (2013). Migrating Mule Deer: Effects of Anthropogenically Altered Landscapes. PlosONE, 8(5), 1-10. 5Bureau of Land Management (2009). Bats of the WRFO. Unpublished report. Accessed Feb 19, 2017. All images licensed by creative commons or captured by Dr. Buxton or Trent Hawkins. Thanks to C. Garsha and M. Mckenna at the Natural Sounds and Night Skies Division of the National Park Service for help with acoustic equipement and P. Lendrum for help in the field. Also, K. Crooks, L. Angeloni, G. Wittemyer, and K. Fristrup for assistance during project development.

●  We used bat call recognition software

SonoBat (version 4) to determine: •  If each triggered recording was actually a

bat call; if the recording was a bat call, what species was it; and the probability that the call is correctly identified to species.

●  We used arcmap (version 10) to determine the distance between each recording site and the nearest road and oil and gas feature.

Dr. Buxton sets up an SM2 at site SME (left), and a map of the study area West of Meeker, CO (right)

Preliminary Results ●  Well distance (m) and road distance (m) are negatively

correlated with the number of bat calls detected at each site over the course of the entire recording period.

SonoBat software used to identify bat calls

Graphs of Well Distance (R2= 0.0325) (top-left) and Road distance (R2=0.0356) (top-right) vs Number of Calls. The bottom-center graph represents Amount of Canopy Cover vs. Total Bat Calls (p-value=0.8884).

●  Devices were set to record 1.5 hours after sunset, and 5 minutes each 30 minute period until sunrise. Recordings were automatically triggered during this time period by any sound >20 kHz.

●  Habitat surveys were performed at each site (canopy cover, canopy height, tree species, scrub cover)

?

●  Linear model results were unexpected - as distance from infrastructure increases, the number of bat calls decreases.

●  It is possible that these trends could be explained by the amount of canopy cover, where we found more bat activity at sites with more tree cover.

●  Trends could also be explained by species composition, where the most common species found in our recordings was the Silver-haired Bat (Lasionycteris nocitivagans)