SHIPBOARD SURVEYS OF MARINE BIRDS 



observers who have other duties during the cruise and only 

 limited observation time. 



Transects taken continuously for an extended period of 

 time are used while a ship is moving between two points, 

 or when conducting radials either perpendicular to a given 

 location (such as a breeding colony) or parallel to it (along 

 or across a habitat such as the ice edge, fronts, or con- 

 tinental shelfbreak). If only a single observer is available, 

 short breaks should be inserted in the series at predeter- 

 mined intervals. If two or more observers are available, 

 they should alternate recording transect data (hourly), or 

 one should scan the transect while the other transcribes 

 data and occasionally relieves the first observer. 



With dedicated ship time, the observer can take sets of 

 transects within each identifiable habitat in the study area. 

 Unfortunately, our knowledge of the oceanography of 

 most areas is limited and it usually takes a considerable 

 amount of sampling to identify and define oceanic habitats. 

 Kessel (1979) has classified major habitat types for Alaska 

 and Favorite et al. (1976) describe the marine environ- 

 ment of the subarctic Pacific in terms of domains and cur- 

 rent systems. There are a few fairly reliable clues that 

 can be used to identify habitats, such as major or rapid 

 changes in surface water temperature, depth, or salinity. 

 The problem of adequate sample sizes for habitats has still 

 not been resolved for our techniques. Seabirds are fre- 

 quently clumped, even within apparently homogeneous 

 habitats, and their distribution pattern may change 

 dramatically in a relatively short time: a density index of 

 1 bird per square kilometer at 0800 h in a given location 

 may change to 1,000 birds per square kilometer in the 

 same location at 0900 h. This is particularly true for 

 species, that may occur in very high densities and tend 

 to form enormous, short-term aggregations from many 

 small, wandering flocks (e.g., the short-tailed shearwater, 

 Puffinus tenuirostris). The number of transects needed to 

 adequately form a mean density index for a given area 

 will vary depending on the distribution of birds, the 

 homogeneity of the habitat, and the extent of the habitat. 

 Sample sizes should be as large as possible. 



Seabirds are not uniformly or predictably active 

 throughout the day, and different species almost certain- 

 ly have different activity cycles. This variation should be 

 recognized and allowed for by conducting surveys dur- 

 ing as many parts of the day as possible. 



Bays 



Most bays and passes have varied bottom topographies, 

 substrates, and tidal conditions. These factors affect the 

 distribution of marine birds and their foods, and dramatic 

 changes in abundance often occur over short distances and 



times. Transect paths within these habitats should sam- 

 ple the varied bottom topography; for example, zig- 

 zagging from shore to shore across the area. Sample as 

 many of the habitats available to marine birds as possi- 

 ble, hopefully sampling each habitat in proportion to its 

 availability to the birds. Shoreline habitats are usually 

 undersampled when following the zig-zag pattern sug- 

 gested above. When possible, adjustments should be made 

 to the cruise tracks in order to bring the percentage of 

 coastline sampled closer to the percentage of other habitats 

 sampled. Timing of transect coverage is more important 

 in bays than it is in pelagic areas; tides, for example, have 

 a great effect on seabird activity within bays and passes. 



Observation Techniques for 

 Strip Transects 



Each survey unit (e.g. , 10-min counting period) is called 

 a transect. The width and length of the transect define a 

 rectangle; the area within the rectangle is the count zone 

 of the transect (Fig. 1). Determining whether a bird is 

 or is not counted depends on how a transect is defined 

 and on the location and movement patterns of the birds. 

 We recognize three basic types of flight patterns for this 

 purpose: (1) feeding flight is when the bird actively 



O-G- 





LEGEND 



Traveling bird not counted O Feeding or foraging bird not counted 

 Feeding or foraging bird counted 



Traveling bird counted 

 Instantaneous counts 

 Transect area 



Bird previously counted 

 Ship 



Fig. 1. Bird sightings to be included or excluded from transect, 

 with three instantaneous counts of flying birds. 



