A04_Statement_of_Work_CHIS_Monitoring_of_Seabirds.pdf

Monitoring of Crevice Nesting Seabirds at Channel Islands National Park using Automated Recording Units and Machine Learning

Channel Islands National Park Solicitation #140P8425Q0076

Contact: Stacey Ostermann-Kelm, MEDN Program Manager and Ecologist, 1 Baxter Drive, Suite 180, Thousand Oaks, CA 91360. Phone: 805-338-3955. Email:

stacey_ostermann@nps.gov

Project Need:

Knowledge of the distribution of crevice nesting seabirds is critical to allow Channel Islands National Park (CHIS) managers to make science-based decisions regarding opening trails or other recreational opportunities in seabird habitat. Knowledge of the distribution of seabirds within the park is also necessary to determine an appropriate sampling approach for long-term monitoring of seabirds at CHIS.

Project Description:

The Mediterranean Coast Network (MEDN) of the NPS Inventory and Monitoring (l&M) Program conducts long-term ecological monitoring within national park units in southern California. CHIS supports globally significant seabird populations. Seabirds at CHIS were identified as a high priority "Vital Sign" for the MEDN long-term monitoring program.

CHIS has been monitoring seabirds as part of the I&M program since the late 1980s. A review of the MEDN I&M Program conducted in 2022 recommended updating our seabird monitoring protocol. Recent advances in technology and artificial intelligence are making it increasingly cost-effective to monitor multiple species of nocturnal, secretive, crevice nesting seabirds. Recordings of seabird calls collected by autonomous recording units (ARUs) allow monitoring to occur for months at a time with minimal human involvement. Advances in machine learning and have made it possible to process thousands of hours of field recordings to identify and quantify acoustic signals from target species. The combination of passive acoustic sensors and automated signal detection is a powerful and cost-effective technique for monitoring elusive species at CHIS such as Ashy Storm Petrels (ASSP), Scripps’s Murrelets (SCMU), and Cassin’s Auklets (CAAU), and to screen for the presence of Leach’s Storm Petrels (LESP), Guadalupe Murrelets (GUMU), and Rhinoceros Auklets (RHAU).

For this project, we are seeking assistance with the design of the inventory, sensor deployment, and the analysis of 500 ARU detector months (approximately 36 hours of data per month per detector) collected in 2025-2029 (see Table 1 for annual estimates) for the following species: CAAU, SCMU, and ASSP. For a subset of sites, we also would like to identify LESP, GUMU, and RHAU. Pending successful completion of the proposed work, and availability of additional funding, additional work may be added.

mailto:stacey_ostermann@nps.gov

Recordings will be between February and August, depending on the target species.

GPS coordinates and photos will be taken during each deployment. Prior to placing the unit, the following information will be recorded: name and organization, location, date and time, ARU unit number/name, current weather conditions, and species present and anticipated. This recording ensures that metadata are archived with acoustic data. Units will be programmed to record one minute of audio every ten minutes from dusk to dawn, following protocol the recommended by Conservation Metrics (M. McKown, personal communication).

As part of the pilot phase of the inventory, two SongMeter4 ARUs (Wildlife Acoustics) were deployed at Santa Barbara Island to monitor Scripps’s Murrelets and other potential crevice nesting seabirds in April-August 2025. In 2026 and 2027, MEDN and CHIS staff, aided by collaborators, will conduct an inventory to determine the distribution of crevice nesting seabirds breeding within the park. Approximately 25 SongMeter 4 ARUs will be deployed across the five park islands to determine the distribution of Ashy Storm Petrels (ASSP), Scripps’s Murrelets (SCMU), and Cassin’s Auklets (CAAU).

Some locations will also be screened for the presence of Leach’s Storm Petrels (LESP), Guadalupe Murrelets (GUMU), and Rhinoceros Auklets (RHAU). We realize the need to improve the existing machine learning models for these three species will be an added cost.

We also have data from past ARU deployments that may be prioritized for analysis rather than future deployments. The NPS Seabird Protocol Development Team will work with the contractor to prioritize past and future data for analysis.

Table 1. Summary of planned ARU deployment to monitor crevice nesting seabirds 2025-2029. Adjustments to the number of ARUs and dates deployed may be needed.

Year

No.

ARUs deployed

Months of

Deployment

Total Detector Months

Notes 2025* 2 April - August 10 Pilot Phase

2026 25 February - August 175 Inventory Phase

2027 25 February - August 175 Inventory Phase

2028 10 February - August 70 Monitoring Phase. Could be increased to 175 detector months, pending additional funding.

2029 10 February - August 70 Monitoring Phase. Could be increased to 175 detector months, pending additional funding.

*Note: work during 2025 will involve starting the development of models for LESP, GUMU, and RHAU as well as analysis of data collected on Santa Barbara Island.

Project Objectives and Deliverables:

(1) Advise on the design of the inventory and detector deployment.

(2) Develop/refine machine learning models for LESP, GUMU, and RHAU.

(3) Analyze 500 detector months of ARU data for the presence of ASSP, SCMU, CAAU. A subset of these deployments will also need to be screened for LESP, GUMU, and RHAU.

(4) Write summary reports for each year that summarize call rates per species.

(4) Make recommendations to improve program efficiency and effectiveness.

Results will be provided in the form of annual reports, in addition to conference calls and webinars with NPS staff, resource managers, and collaborating scientists.