3.3 Aerial Surveys (Harlequin Ducks)

Recommendations: Fixed-wing aerial surveys are recommended to determine presence over large scale marine areas or census wintering and moulting populations of harlequin ducks over large geographic areas in conjunction with surveys for other waterbirds. Helicopter aerial surveys are recommended for inventory of breeding pairs and broods on rivers with good visibility. Helicopter aerial surveys can be used to determine presence and to provide relative abundance estimates. On streams with good visibility, and using a helicopter, estimates of absolute abundance can be made.

Fixed-wing aerial surveys have been used extensively for censusing general waterbird populations at sea, in estuaries, and on breeding wetlands in British Columbia (e.g., Savard 1979, 1982), but few attempts have been made to survey streams. In coastal British Columbia, few attempts have been made to census moulting or wintering harlequin populations by air (Savard 1988). Aerial surveys of harlequins on the coast are hampered by the cryptic plumage of moulting males, females and juveniles, and by their behaviour of hauling out on rocky islets or swimming among rocks, which makes them difficult to spot at high speeds (Savard 1988). They can not be adequately detected by flying transects but require passes over suitable islets or along shorelines.

In Labrador and Quebec, helicopter aerial surveys have been conducted specifically, and successfully, for breeding harlequins along rivers (Goudie 1988; Goudie et al. 1994b; Morneau et al. in prep.). Similar surveys have been used successfully in Alaska where 750 km where surveyed by helicopter (McCaffery 1996; McCaffery and Harwood 1996). However, in high density areas ground truthing indicated that up to 50 % of the birds were missed. On breeding streams, aerial surveys are hampered by tree cover, terrain, vegetation, observer bias, and water hydraulics which all can reduce visibility of harlequins. Helicopter surveys can be especially effective in subalpine areas because of better visibility (I. Goudie pers. comm.).

Two breeding season helicopter aerial surveys have been conducted in British Columbia. The first was a survey of Carbon Creek in May 1977, which was successful in finding breeding pairs (10 pairs over 19.5 km; Breault and Savard 1991). This method was tested in June 1996 on the Nahatlatch River (Boston Bar) and proved to be ineffective due to high tree height which forced the helicopter to fly above the optimal height, and riparian vegetation which obscured birds. Detection was estimated to be 30% (n=4) of ground-detected birds.

Clearly, surveys by fixed-wing aircraft for wintering or moulting harlequins present a few difficulties, notably low detection rates associated with higher speeds and heights. Helicopter surveys have several advantages such as they permit low level flying (15-20 m if tree cover permits), slower speeds, and high maneuverability, but are relatively expensive. On wintering and moulting grounds, helicopter surveys of rocky islets and shorelines are also feasible for the same reasons. Although no estimates of the reliability of aerial surveys for harlequins have been made, it seems apparent that helicopter surveys provide satisfactory data.

Aerial surveys often have a consistent bias, the underestimation of population densities (Pollock and Kendall 1987). Comparisons of aerial surveys with ground-truthing or boat surveys are recommended to calibrate abundance estimates. Furthermore, for species such as coastal sea ducks, errors may be random and cannot be adjusted by standardized correction factors (see Savard 1982).

Video cameras and aerial photography are sometimes used in conjunction with aerial surveys. Aerial photography could improve estimates of wintering or moulting flocks of harlequins, as well as provide valuable information about habitat, though this technique has not be attempted.

Advantages

• Helicopter surveys are effective in marine areas and along streams that are safe to fly and have good visibility.
• Large areas can be covered in a short period of time.
• Useful for reconnaissance of potential breeding areas, moulting areas, or wintering areas.
• Remote areas can be easily accessed (i.e., alpine areas, headwaters of rivers, unnavigable sections of rivers, offshore islets, remote coastlines), especially if surveys have to be repeated at regular intervals.
• Since counts are relatively instantaneous, biases caused by movement of harlequins between areas can be avoided.

Disadvantages

• Suitable only for harlequins.
• Costly method that is feasible only when considering high priority species.
• Very difficult and thus inappropriate to survey forested streams because of reduced visibility and safety considerations.
• Requires low-level flying especially on breeding streams (possibly hazardous).
• Wintering and moulting harlequins are difficult to detect from the air.
• Aerial surveys invariably underestimate populations because birds will be inevitably missed. Harlequins prefer to stay close to shore or haul-out on rocks and are difficult to detect. Ground-truthing is likely necessary to correct for these biases but must be co-ordinated with aerial surveys.
• Variables such as light conditions, different observers, altitude, and precise route flown tend to affect the estimate of bird numbers. In some cases these difficulties lead to random error which cannot be adjusted by a standard correction factor (Stott and Olson 1972). Higher quality personnel will, hopefully, minimize error.
• The sight and sound of the aircraft may disturb birds and cause some birds to flush ahead of flights and consequently be missed in counts or, if detected, be double counted at another site.

Aerial surveys are, essentially, encounter transects that follow the contours of shorelines and inland water courses.

Office Procedures

• Review the section, Conducting a Wildlife Inventory, in the Species Inventory Fundamentals (No. 1).
• Obtain relevant maps for project area (e.g. Nautical charts, 1:50 000 air photo maps, 1:20 000 forest cover maps, 1:20 000 TRIM maps, 1:50 000 NTS topographic maps).
• Determine Biogeoclimatic zones and subzones, Ecoregion, Ecosection, and Broad Ecosystem Units for project and study areas from maps.
• Select study areas from 1:20,000 and 1:50,000 maps on the basis of personal experience and existing data.
• Survey routes must be chosen to maximize the coverage of suitable habitat.
• The survey route should follow a transect along coastal shoreline or stream courses and should be drawn on maps used for navigation. The survey route is traced lightly on a 1:5 000 map.
• Contract a suitable aircraft and pilot with previous experience in wildlife surveys.
• The desired survey altitude and flight speed are discussed in advance with the pilot. These will vary with the type of aircraft and weather conditions and determine, in part, the detail at which the data are collected.
• Along the coast, tidal cycles must be calculated for the area to be surveyed and coordinated with the flight plan. High tides are usually best.
• Authority to fly over sensitive areas (Ecological Reserves, Wildlife Management Areas), must be obtained in advance of the survey.
• Maps, recording equipment, and recording forms should be assembled and double checked at least a week prior to the survey date. This will allow for equipment checking, repair, replacement, and photocopying of forms.
• Prior to the survey, personnel should meet to standardize recording and spotting estimates. Team members should review maps of the area.
• Data are recorded on cassette tape during the flight if observations are frequent and transcribed to data sheets as soon as possible after the flights. Tape transcription time should be scheduled for the survey team shortly after the flight. This will reduce errors and allow changes in procedures for more efficient recording on following flights. If observations are infrequent, data may be recorded directly on maps and accompanying data sheets. Harlequin breeding densities rarely exceed 1 pair/km of river.

Sampling Design

• Systematic. Transects follow the shoreline along the coast or stream courses. The entire survey route may be broken up into smaller segments to facilitate counting.

Sampling Effort

• If only one survey is to be scheduled, it should coincide with the peak in moulting or winter abundances, as determined from existing data. However, in most cases, surveys should be repeated several times during peak abundances.
• The time required for a survey depends on: (a) distance covered, (b) level of intensity of the survey and detail of information to be collected, (c) travel time to and from the survey location, (d) number of times the survey is to be repeated, and (e) fuel limitations.