3.3 Presence/Not detected and Relative Abundance

Recommended method(s): Point counts at night with audio playback.

3.3.1 Point Counts

The point count method is the most efficient and data-rich method of monitoring landbirds, and it is probably the most common census method (Ralph et al. 1993). The North American Breeding Bird Survey is a well known study that uses this method. Although no single standardized point count protocol exists, a detailed description of this method can be found in Hilden et al. (1991).

In its simplest form, the point count consists of a single observer visiting an established network of points or a series of points along a predetermined route (e.g., road, trail, railway line). At each point, the observer spends a set period of time, and looks and listens for all birds in the area, noting what was detected. If counts are done within an arbitrary fixed radius, or if distance-to-bird estimates are used, then relative abundance can be estimated (e.g., Buckland 1987; Bibby and Charlton 1991; Hilden et al. 1991). Points are usually revisited at subsequent dates to generate data on population trends over time.

Point counts are an attractive method because they are simple, efficient, data-rich, easily repeatable, and they require little equipment and few resources. The greatest limitation to point counts is the competence of the observer. A great deal of expertise is required to census large bird communities, and the observer must be highly skilled at both visual and vocal recognition. Another shortcoming occurs if birds approach or flee from the observer before he/she has reached the census point. This is especially common in open areas, such as prairies or meadows. In these locales, line transect methods may be preferable because they can account for fleeing birds. The assumptions of the point count method are as follows (Bibby et al. 1992):

0. do not approach the observer or flee;
0. are 100% detectable at the observer;
0. do not move much during the count period;
0. behave independently of one another;
0. of the above assumptions do not interact with habitat or elements of study design;
0. estimates are accurate; and
0. are fully and correctly identified.

Point counts are the easiest and most productive method for censusing nocturnal birds, such as goatsuckers. This method is recommended above other methods because of its simplicity and efficiency. Point counts for goatsuckers differ from normal methods in a number of ways. The foremost difference is that the census period must be in the evening rather than the morning.

The drawback, as for all activities involving nocturnal birds, is that sampling must occur during twilight hours or at night, not a traditional time for most ornithologists to be active. To our knowledge, in North America there is only one published account of a method specifically for a poorwill census (Kalcounis et al. 1992) and nothing exists for nighthawks. Kalcounis et al. (1992) used point counts along existing roads, trails, and footpaths to census poorwill distribution and abundance in the Cypress Hills region of Saskatchewan. A similar protocol could be applied to nighthawks.

Point counts also appear to be the most common method for other Caprimulgid species. Kepler and Kepler (1973) used roadside point counts in conjunction with call playbacks to assess distribution of the Puerto Rican Nightjar (Caprimulgus noctitherus) in Puerto Rico. Two similar studies have been conducted by Vilella and Zwank (1987, 1993). Cooper (1981) used roadside point counts to assess the distribution and relative abundance of two Caprimulgid species (Whip-poorwills (Caprimulgus vociferus) and Chuck-will's-widows (C. carolinensis) in Georgia, U.S.A. In addition, point counts and small scale area searches have been used in Great Britain to determine the distribution and relative abundance of European Nightjars (1989). In all cases, point counts have relied on listening for calls. Although a drastic change to this protocol is not recommended, the usefulness of "night-lighting" for eyeshine to enhance the method needs to be evaluated.

Call Playbacks

During the breeding season, male poorwills and nighthawks are territorial, and will respond aggressively to another male that is calling within its territory. Normally, the playback of recorded calls using a portable tape player will elicit a response from these birds, and thus, will increase the chance of finding birds.

This is a useful technique to determine the total number of males within a relatively small area. Assuming that the sex ratio is 1:1, and that mates have already been found, the total number of adult birds in the area can be approximated by multiplying the total number of males by two (e.g., Kalcounis et al. 1992). It should be noted that this type of censusing should be conducted only during the early breeding season when calling activity is at its greatest. At other times in the summer, birds may be less likely to elicit a response to playbacks which may lead to inaccurate and highly variable population estimates (see McNichol 1981). This is especially true for male poorwills, who apparently suspend calling, late in the breeding season (R.D. Csada, pers. comm.; D. J. Bender, pers. obs.).

Office Procedures

• the introductory manual, Species Inventory Fundamentals (No. 1).
• relevant maps for study area (e.g., 1:50 000 air photo maps, 1:20 000 forest cover maps, 1:20 000 TRIM maps, 1:50 000 NTS topographic maps). 1:50 000 scale maps from the Canadian topographic map series provide the locations of roads, streams, human-made structures, lakes, the presence of vegetative cover, etc., and they can be used to interpret the layout and topography of the landscape. Aerial photographs or fine-grained remote sensing images can be used to obtain more precise details such as the location of trails and footpaths, or the composition of the vegetation cover.
• on map interpretation, identify routes (transects) for census.
• Biogeoclimatic zones and subzones, Ecoregion, Ecosection, and Broad Ecosystem Units for census areas from maps.

Sampling Design

• type of survey can be conducted either by foot, or if roadways exists, by vehicle. The latter is more efficient because less time is spent on travel between points, and more time can be spent observing and collecting data. Under optimal conditions, 25 to 30 five-minute point counts can be conducted in three hours using a vehicle. Surveys by foot may be preferable in areas of rough terrain where vehicle travel is not possible and because vehicle surveys tend to be biased against species that avoid roads. However, surveys by foot are more time consuming, and depending on the terrain, generally only one quarter to one half as many points are possible.
• size of the survey area will depend on the level of monitoring desired, as well as the people-power available to survey the area. Once the area of the survey has been established, individual transects for the census can be set. The best routes are those that are easily recognizable, especially in the dark, such as roads, trails, or foot-paths.
• transects are to be revisited, each point on the transect must be identifiable, again especially by night. When possible, permanent markers should be established to indicate the points. If this is not possible, as would be the case within a provincial or national park, then points may have to be selected at landmarks along the transect that are distinct enough to be recognized at night (e.g., road intersection, signpost, bridge, etc.).
• choosing a census route:
• existing roads, trails, footpaths, etc.
• tertiary roads first, then secondary roads, and avoid well travelled, wide roads. Surveys conducted along less travelled roads frequently yield data that are most representative of the natural habitat because some birds may avoid (or prefer) high traffic roads.
• transects are surveyed by vehicle, points should be sampled about every 500 m. Assuming that an observer can cover approximately 15 five-minute stops in an evening, transects should be about 7 to 8 km in length.
• travelled by foot should have sample points spaced at least 400 m apart. Assuming that an observer can cover approximately seven five-minute stops in an evening plus travel time, realistic transects will be only 3 to 4 km in length.
• relative abundance estimates a fixed radius or distance-to-bird method should be used. Distance to birds may be difficult to obtain in the dark, particular for flying nighthawks.

Sampling Effort

• Time requirements for censusing are determined by the extent of the census.
• Assume that only one transect can be conducted per night. Therefore, a minimum of 2 person·hours/night are required for each transect in the census.
• The time over which censusing can be conducted is limited by the length of the crepuscular period. Generally speaking, in May and June, the entire crepuscular period is only about 1.5 hours.
• Repeat census several times during early breeding season.

Personnel

• Minimum personnel required is just one person, although two are preferable for off-road work conducted at night.
• Personnel must be familiar with call identification for nighthawks and poorwills. This skill may be learned by listening to recorded calls. Once heard, nighthawk and poorwill calls are unmistakable and easily recognized.
• surveys must be conducted at night by personnel familiar with night navigation.
• must be very familiar with the use of a map and compass.
• possible, sites should be visited during daylight hours so personnel can familiarize themselves with the route.
• with precautions, personnel will get lost on occasion; thus, the need for at least two individuals for surveying remote locations. It is also helpful to have two-way radios to enhance safety.