3.1 Sampling Standards

The following are guidelines for conducting standardized woodpecker inventory studies in the province. Close adherence to these guidelines will permit the collection of reliable data that should satisfy individual and corporate inventory needs, as well as contribute to biodiversity monitoring at local, regional, and provincial scales.

3.1.1 Personnel

• The number of surveyors required will depend primarily on the size of the sampling area and the study objectives. For areas less than 20 ha, two surveyors may be sufficient to cover the area during each sampling period. For larger areas, teams of four or six crew members may have to be employed.
• Crew leaders should be professional biologists with woodpecker experience. Crew members should either have experience in woodpecker surveys or receive proper training prior to the sampling period.
• Useful field training tools are the "Guide to the Vocalizations of Woodpeckers in North-eastern Oregon", a cassette produced by the U.S.D.A. Forest Service (1992) and the "Training Guide for Bird Identification in Pacific Northwest Douglas-Fir Forests" (U.S.D.A. Forest Service 1990). A detailed procedure for training observers is given in Kepler and Scott (1981).
• Crew members should be in good physical condition and have comparable surveying ability.
• Crew Leaders or the Project Biologist must have working knowledge of the statistical concepts of surveys as presented in the RIC Species Inventory Fundamentals manual and associated references.

3.1.2 Controlling for observer bias

Various studies have shown that observer bias is one of the most noteworthy bias factors in trend analysis of many bird populations. In fact, one study suggests that a potential reason for the apparent recent increase in some songbird populations (as determined by breeding bird surveys) is the apparent increase in skill of birdwatchers (Sauer et al. 1994). In another study, it was found that a significant change in trend resulted if individual observer's first year of observation in a breeding bird survey was removed (Kendall et al. 1996). However, if observers are trained appropriately from the start of a project, methods to account for observer bias should not be required (see note 5 below). Strategies to avoid observer bias include:

1. When possible, change observers between stations, transects, or rotate observers between habitat types on repeated surveys to minimize recurring bias in any segment of a survey;
2. Observers should be tested to ensure that bias due to misidentification of calls or drumming is minimized. The performance of observers should be recorded for possible use as a weighting factor or criteria for stratification in the analysis of data as described in Point 5 (below). Obviously, these tests should be done prior to surveys;
3. Field coordinators must ensure that observers include all woodpecker observations on appropriate data sheets using standard species codes as designated in this manual. Woodpeckers that were observed either visually or aurally, but for various reasons could not be identified to species, should also be recorded on data sheets to the highest taxonomic classification possible (e.g., Picidae, for an unknown woodpecker species). It may also be useful to include a written comment as to why a species could not be identified (e.g., heard calling, but could not locate for positive ID). Particularly in dense forest habitats, it is not always possible to identify each woodpecker observation to species. Crew leaders should ensure that field personnel are aware that, although it is important to make a best effort to identify species (within the constraints of the survey methodology), it is better to document an "unknown" rather than record guesses. Crew leaders must ensure that each observer has a suitable level of competence in identification;
4. Replicating counts from an individual site can identify the influence of within-site variability on results using the methods of Link et al. (1994). Within-site variability can be defined as variation due to factors such as differences between observers, and short-term variation in population size at a count station or monitoring site. This is not to be confused with between-site variability, which is due to large-scale differences in the spatial distribution of species, and forms the basis for most experimental designs. In general, Link et al. (1994) found that if the proportion of within site variation is large, and the cost of replicating a site is small compared to setting up a new site, then it is optimal to replicate counts. If the proportion of within site variation is small, and the cost of replicating a site is equal to that of setting up a new site, then it is optimal to not replicate. Not surprisingly, this study found that counts for birds with lower abundance, such as many woodpeckers, had the highest percentage of within-count variation. Therefore, project biologists should consult Link et al. (1994) when designing monitoring studies, especially for woodpeckers that will have low average counts. In general it will be expensive and time-consuming to locate sufficiently large and contiguous sites for woodpecker surveys, and access to those sites for establishing point count stations or transects will be problematic. Therefore, it is expected that replication of sites will be more cost-effective than establishing new sites;
5. If there was significant variation between observer skills, data sets can be tested for observer effects by stratification by observers (ANOVA) (Buckland et al. 1993) or addition of covariates or weighting factors for trend models (Sauer et al. 1994, Thomas 1996, Link and Sauer 1997). However, this is not necessarily a good strategy for a reduction in the power of tests and precision of estimates may result with the addition of covariates (to trend analysis) or strata (to ANOVA) designs. Power analysis in the design phase can be used to explore this problem (see RIC Species Inventory Fundamentals manual). The best strategy is to use qualified observers or to train new observers adequately to minimize potential bias rather than rely on complex statistical analysis.

3.1.3 Time of day

• Woodpeckers, like many birds, are most active in the morning. Therefore, surveys should be conducted from about one half hour after sunrise until noon.
• Hartwig (1999), found that the most reliable time of day for call playback responses of Pileated Woodpeckers on south-eastern Vancouver Island was from 1.5 to 3 hours after sunrise, although they would still reliably respond at about a 1:10 ratio of response to effort up to 5.5 hours after sunrise.

3.1.4 Time of year

• The survey season depends on the particular study objectives. In general however, surveys are most efficiently conducted during the breeding season (when birds are generally most conspicuous). Timing and length of the breeding season is species-specific (see Section 2) and varies with latitude and elevation. Furthermore, pairs within populations differ in their timing of breeding and annual activity schedules. These differences must be considered for sampling during the breeding season.
• Winter surveys are more likely to be affected by inclement weather and snow pack, both of which may reduce the rate of observer travel (from 1 km/hr in spring to 0.5 km/hr during the winter; Carey 1983). However, the visibility of birds may be enhanced in winter when deciduous trees are leafless. Studies considering assessments of impacts on bird populations may wish to consider resident (i.e., most woodpecker species) as well as migratory populations. In such cases, breeding season surveys as well as winter surveys are warranted and highly recommended.
• Call playback surveys should be conducted before or during the breeding season when birds are most territorial. Other survey methods may be employed throughout the year.

3.1.5 Environmental conditions

Poor weather such as high winds, rain, and fog can inhibit both bird behaviour and observer ability (Table 1). High winds and rain are more of a problem in forests than open grasslands due to increased noise in the canopy.

Table 1. Acceptable and unacceptable weather conditions for woodpecker surveys.

	Acceptable	Unacceptable
Wind	Beaufort 0 (< 2 km/hr). Smoke rises vertically. Beaufort 1 (2-5 km/hr). Some smoke drift. Beaufort 2 (6-11 km/hr). Leaves rustle.	Beaufort 3 (12-19 km/hr). Leaves and twigs in motion. Beaufort 4 (20-29 km/hr). Raises dust - small branches move. Beaufort 5 (30-39 km/hr). Small trees sway. Beaufort 6 (> 40 km/hr).
Precipitation	None Light drizzle Light snow (winter)	Steady rain Heavy snow
Temperature	> 7 0C ( breeding) > 0 0C (winter coast) > -10 0C (winter interior)	< 7 0C (breeding) < 0 0C (winter coast ) < -10 0C (winter interior)

3.1.6 Habitat data standards

Habitat data that has typically been correlated to woodpecker abundance includes stand age (structural stage), DBH, tree species composition (e.g., conifer, mixed-wood, deciduous), stand density, and the number and quality of dead or dying trees. A minimum amount of habitat data must be collected for each survey type. The type and amount of data collected will depend on the scale of the survey, the nature of the focal species, and the objectives of the inventory. As most, provincially-funded wildlife inventory projects deal with terrestrially-based wildlife, standard attributes from the terrestrial Ecosystem Field Form developed jointly by MOF and MELP (1995) will be used. The manual, Species Inventory Fundamentals (No.1), contains a generic discussion of habitat data collection as well as a list of the specific requirements for woodpecker surveys (Appendix E).

3.1.7 Survey Design Hierarchy

Woodpecker surveys follow a sample design hierarchy which is structured similarly to all RIC standards for species inventory. Figure 1 clarifies certain terminology used within this manual (also found in the glossary), and illustrates the appropriate conceptual framework for a call playback survey for woodpeckers. A survey set up following this design will lend itself well to standard methods and RIC data forms.

Figure 1. RIC species inventory survey design hierarchy example.