Storm Petrel Census and Monitoring
Julia Ambigis '01, Master's Thesis
The purpose of my study was to learn more about Leach's Storm Petrels. We need more information to be able to census and monitor petrel population fluctuations, which will allow us to better assess the overall health of these populations. To this end, my study had three main objectives; to estimate the population size of petrels on Great Duck Island, Maine, to compare different ways to census petrels, and to gain a further understanding of petrel habitat preferences.
Great Duck Island off the coast of Maine is a 109-hectare island of mixed open and forested areas. Censuses on this island have reported a range of 5,000 to 16,000 breeding pairs (Table 1). No two surveys have been done the same way and none of the surveys are comparable (Cowger 1976, Folger and Wayne 1986, Huntington 1963).
I sampled 13 open and nine forested plots. This was a total of 55,000m2 or 7.5% of suitable petrel habitat on the island. I used a Shreco Flexible Camera system to determine burrow occupancy. This camera is designed to examine the inside of sewage pipes, in this application the camera was inserted into the burrows. The camera, attached to the end of a flexible cable, was maneuvered with the cable and a thin cord attached to the head of the camera. This presented a picture of the burrow. Roots, cones, nesting material, bird, eggs, and even insects were discernable. A burrow was determined to be active if it contained an adult, an egg, a chick, or an eggshell. I counted each active burrow as one breeding pair.
I found that Great Duck Island, Maine, has 9297.3 + 6500 (90% CI) breeding pairs of Leach's Storm-Petrels.
I compared three techniques for censusing Leach's Storm Petrels. Estimates of Leach's Storm Petrel population densities range widely even from the same island. Difficulties in studying this species are that they nest in burrows, are nocturnal, and spend most of their life out at sea, only coming in to land to breed. I compared traditional "grubbing" (reaching in the burrow with ones arm), to use of a camera on a cable to visually inspect the burrows, and playback recordings to elicit a response from occupied burrows. Traditional grubbing proved to be inefficient and detected the fewest active burrows. Playbacks detected as many burrows as the camera method, took less time and required less equipment. The playback method is less intrusive than the other two methods because it allows detection of a bird with no need