Author: Sam Droege, USGS Patuxent Wildlife Research Center, 12100 Beech Forest Rd., Laurel, MD 20708, frog@usgs.gov, 301-497-5840.

Species list
Description of technique
       Field time components
       Office time
       Equipment
Things that could bias your counts
       Observer skill
       Attractants
       Feeder birds
       Changes in the places within the count circle where bird watchers count birds
       Changes in the number of participants
       Date of count
       Shifts in the count center point
       Changes in transportation
       Weather
Advantages and disadvantages
Approaches to analyzing your data
Existing protocols and programs using this technique
Studies that have used this technique
Literature cited
Send a comment on this technique (this takes you to another page)
See existing comments (this takes you to another page)

All birds wintering in North America.

Christmas Bird Counts (CBC) were first started in the very early years of the 20th Century by the National Audubon Society as the Christmas Bird Census (Anon. 1900) “established primarily as a means of arousing interest in field work and of encouraging definite methods of recording one’s observations” (Anon. 1907). A century later the National Audubon Society continues to coordinate and publish the CBC but now with over 1800 counts completed each year by over 50,000 participants. At its inception, these counts simply required participants to tally the number of species they located on a given day around Christmas time and include a count or estimate of the number of individuals that were detected for each species. The counting area was defined only to the nearest town, multiple observers often repeatedly counted the same birds, and when more than one observer was present, it was usually unclear as to whether they worked together or separately. As early as 1913 the editor was attempting to keep multiple observers from double counting and by 1916 (Anon. 1916) the editor was requiring that a count limit itself to a 15-mile diameter circle. Further notes on the history of the CBC.

Currently, observers are required to track the number of miles they travel and the number of hours they count birds during each CBC. The type of transportation they use (foot, boat, car, etc.), the number of hours feeders were watched, and if they are birding during the day or night are also recorded. Bird data for a CBC circle are printed in the yearly Christmas Bird Count issue of the journal American Birds as a simple list of number of individuals of each species detected on the day of the count with scattered notations regarding rare or unusual species some times added. Data are also posted on the web. Unfortunately, for logistical reasons, no attempt is made to split the total bird count into counts made by different counting techniques(e.g., feeders, nocturnal counting, boat counts).

In some populated sections of the continent the density of Christmas Bird Counts has reached the point that new ones can no longer be established. In the policy set by the National Audubon Society, new circles cannot overlap existing circles, consequently in many of the more urbanized parts of the country there no longer is room for the creation of additional circles.

As a technique the CBC can be thought of as a primitive form of time- and area-constrained search. In this case, time is restricted simply to a maximum of 24-hours and the area to a maximum of 15-mile diameter circle. As will be shown in the sections below there are many constraints to the use of these data for monitoring. The CBC’s current relevance is as a comparative historical source of information on bird changes, a coarse means of capturing large bird changes, and a conservation-oriented recreational pursuit by birders. There are better ways of measuring changes in wintering bird populations. On the other hand, since the CBC system is already in place at no cost to anyone but those who participate and over long periods of time interesting trends are documented, the running of CBC’s by birders should not be discouraged.

In fact, by collecting data in just a little bit of a different way all CBC’s could be collecting data of far greater value to managers and conservationists while still meeting all the current requirements of the National Audubon Society. While the National Audubon Society has no need themselves to collect these more detailed data sets, a local count compiler could ask their to record the additional information in a way that will greatly increase their value.

If you are considering augmenting your CBC or starting a more rigorous winter survey for birds then you might consider creating a system of permanent counting territories within your CBC. These areas would be small enough that one person or party could cover them in a morning. People covering these areas would be asked to record and keep the data for these territories separately along with information about the amount of time and effort they spend in each counting territory. Coverage outside of these counting territories in the afternoon (when detection rates for winter birds plummet) and in areas not part of these defined territories can be pooled together, as usual. By tracking these small permanent territories separately you can better account for variation in the amount of effort, expertise, and coverage that occurs within a circle. Changes and trends within each of the territories can be investigated and associated with changes in habitat or other factors.

Field time is a simple function of the number of hours individuals participate on the day of the count and approximately another 15 minutes for each participant to transcribe their field lists to a the final form dictated by the individual count organizer or compiler.

The amount of time it takes to organize a CBC count and compile all the data from the participants into a final tally varies with the size of the count and the compiler’s organizational approach. Usually a list of past participants is maintained in a database and a letter/email sent out to participants a month or so before the date of the count asking if they will participate again and where they would like to count within the circle. There can be several hours of correspondence with observers spent recruiting, taking care of problems, and responding to requests. At the end of the day of the count most CBCs hold a 1-3 hour tally of the day’s data with participants. Compiling, dunning those late in sending in their completed information, summarizing the data, double-checking, tracking down errors, and filling out the forms for National Audubon Society can take an additional 5-15 hours.

• Paper
• Pencil
• Clipboard
• Binoculars

Additional equipment used by some individuals or count compilers can include spotting scopes, tape recorders, boats, and a computer for ease of compiling data, however, none are absolutely necessary.

Things that could bias your counts

Traditionally, CBCs welcome any person that has an interest in birds to participate. The result is that the final tally of birds on a CBC counts is the product of the efforts of skilled and unskilled observers. Changes in the ratio of skilled to unskilled participants can have a huge effect on both the number of species recorded and the number of individuals detected as unskilled observers often record far fewer species and individuals than the skilled observer. As each individual’s contribution is hidden within the amalgam of the final count, there is no way to correct each year’s results for changes in the mix of skill levels of participants except in the most anecdotal of ways.

Christmas Bird Count participants are free to use whatever means of attracting birds they desire as they count birds. This includes imitating owls, spishing (spishing is difficult to describe with words, it really must be demonstrated to best understand, but it involves the making of a set of pishing, shushing, spishing, and shooking sorts of sounds that are thought to mimic the alarm calls of some species), or using tape recordings of the species they want to attract (e.g., rails and owls).

Because attractants can increase the number of birds that are detected by an observer, changes in the use of attractants across the years a CBC is run will artificially inflate or deflate the number of birds detected.

In most CBCs a number of participants do not leave their house, but simply count the number and species of birds coming to their bird feeders. While the number of feeder watcher participants and their hours are recorded separately, the data contributed by feeder watchers are pooled with all the other bird data from the count. Additionally, feeder watching participants are confronted with the problem of trying to determine how many birds are coming to their feeders. Guidance is usually not provided, thus some participants record the maximum number of a species they detect at one time while others estimate what they think is the overall number in area that might be coming in throughout the day. Thus counts of seed-eating bird species known to be associated with bird feeders can become greatly biased by changes in the number of participants who only watch birds at their feeders, how those feeder watchers count their birds, and by changes in the type and extent of the feeders used.

Participants are free to count birds wherever they wish within the 15-mile diameter count circle. Changes in the accessibility of areas that contain high densities of individuals or unusual species will bias the results. Landowners, corporations, or government agencies can control access to these “good” birding spots and changes in access are not infrequent. Depending on the uniqueness of these properties, in terms of the number and kinds of birds found there, their accessibility or inaccessibility can bias the results. Again, as with many of the other biases associated with the current CBC methodology, if count data were tracked separately for these special areas, these biases could be removed during analysis.

The mindset of CBC participants and organizers is such that their primary objective remains to count as many species as possible in their CBC circle on the day of the count. Thus the types and extent of coverage any region within a CBC circle garners is usually a function of the perceived probability that that area will help add additional species. This often means that rare and unusual habitats are covered at a disproportionately higher rate than the ordinary. In particular, water bodies and wetlands are usually intensely covered and urban and suburban residential and commercial areas covered only lightly.

Over time, most CBCs increase in their number of participants. When participants are few, much of the coverage focuses on the “best” habitats. As numbers of participants increase, coverage shifts into habitats not thought to yield as many species simply because most of the prime areas are already covered. Consequently, some groups of uncommon birds, such as waterfowl or roosting birds are counted at the same places each year even if participation is low, as they are considered key to achieving a high species count. Consequently, the counts for these species are likely to be less tied to the overall number of CBC participants or the numbers of hours participants spent in the field. In contrast, other, more mundane species, such as most of landbirds, are counted in some proportion to the number of hours participants spend in the field. However, at extremely high densities of observers, the number of birds counted per hour as well as the types of species will change as participants are crowded into less and less favored habitats.

Currently, the National Audubon Society allows a 23 day time period between December 14th and January 5th in which any CBC count can be run. The difference in weather between the first part of the period and the last part can, in some of the middle latitudes, mean moving from above to below freezing temperatures resulting in water bodies freezing, waterfowl moving south, and more temperate-minded birds leaving. As a consequence, if count organizers shift the count from the early part of the season to later or visa-versa, the number and types of species detected can be affected.

Quite a number of Christmas Bird Count circles have moved their center points at some time in their history. The reason that most have done so has been to increase their species count through the capturing of new and better bird watching habitats. This seems like a natural thing to do from the perspective of increasing the overall species tally for the count, but it also effectively creates two circles. Thus most meaningful comparisons between the two circles end when a center point is shifted.

The number, rate, and type of birds seen while driving around in a car differ from those seen from boats or while walking. Participants in CBCs are free to use any mode of transportation they like as they count birds. While the number of hours and miles that each mode of transportation is used is recorded, the resulting bird data are all pooled together in the published accounts. The consequential results are that there is no easy way to account for changes in the ratios and types of transportation over time, creating yet another form of bias within the CBC. Again, if data were kept separately for each type of activity, it would be possible to at least partially accommodate these biases.

Most CBC counts fix the date of their count well in advance so that participants can plan their busy holiday season. A consequence of fixing the date is that in some years the weather can be inhospitable to birds and bird counters. High winds, wet weather, and extreme cold decrease the rate at which birds are detected. Additionally, if weather is bad observers spend more time in their car drinking coffee and staring out their windshield rather than walking around, further decreasing rates of detection.

Advantages:

The advantages are that historical data on some CBCs can go back 100 years, many CBCs already exist, the technique attracts wide participation by birdwatchers, and the technique can easily be changed to diminish many of the existing biases while maintaining both observer interest and the required data for the National Audubon Societies’ publications and database.

Disadvantages:

The technique, in its current form, is greatly compromised by the lack of any means of correcting for differences in observer skill, ways of attracting birds, time-of-day, mode of transportation, and circle coverage. Most of these negatives can be corrected for as outlined in the Introduction section.

Great care must be taken in interpreting the results of CBC counts. As listed above the biases are severe. Researchers attempting to interpret the results of individual CBC counts need to have a good understanding of the history of the count, who participated, how those participants behaved, and what approaches and circle coverages changed during the counts history. At minimum, data must be standardized by divided by the number of hours of observer participation on that count. Anyone interested in using CBCs to monitor birds is urged to at least contemplate the adoption of the simple changes in approach listed in the Introduction and bias sections above.

Examples of analysis approaches can be found in the following publications and web sites:

The Audubon Society Maintains a CBC bibliography including articles about the analysis of CBC data.

Below are some additional references not listed on the Audubon’s site:

Butcher, G.S., M.R. Fuller, L.S. McAllister, and P.H. Geissler. 1990. An evaluation of the Christmas Bird Count for monitoring population trends of selected species. Wildlife Society Bulletin 18:129-134.

Droege, S. 1998. Birds and landscape changes in Northeastern forests. Pages 187-188 In M.J. Mac, P.A. Opler, C.E. Puckett Haecker, and P.D. Doran, editors. Status and trends of the nation's biological resources. Volume 1. U.S. Department of the Interior, U.S. Geological Survey, Reston, VA.

Dunn, E.H. and J.R. Sauer. 1997. Monitoring Canadian bird populations with winter counts. Pages 49-55 In E.H. Dunn, M.D. Cadman, and J.B. Falls, editors. Monitoring bird populations: The Canadian experience. Canadian Wildlife Service, Ottawa, Ontario.

Floyd, T. 2001. A new method for time series analysis of long-term bird population data. Bird Populations Journal 5:1-10.

Link, W.A. and J.R. Sauer. 1999. Controlling for varying effort in count surveys --an analysis of Christmas Bird Count Data. Journal of Agricultural, Biological, and Environmental Statistics 4:116-125.

Link, W.A. and J.R. Sauer. 1999. On the importance of controlling for effort in analysis of count survey data: Modeling population change from Christmas Bird Count data. Pages 119-124 in A.J. Helbig and M. Flade, editors. Bird numbers 1998: where monitoring and ecological research meet: proceedings of the 14th International Conference of the European Bird Census Council (EBCC) in Cottbus (Brandenburg), Germany, 23-31 March 1998. AULU-Verlag, Wiebelsheim, Germany.

Sauer, J.R. and W.A. Link. 2002. Using Christmas Bird Count data in analysis of population change. American Birds: The 102nd Christmas Bird Count :10-14.

Sauer, J.R., S. Orsillo, and B.G. Peterjohn. 1994. Population status and trends of grouse and prairie-chickens from the North American Breeding Bird Survey and Christmas Bird Count. Transactions of the North American Wildlife and Natural Resources Conference 59:439-448.

See also:

• CBC analysis web site from Patuxent for maps, trends, and analyses
• Forest Bird Monitoring in Fundy National Park
• Refining the Data Collected During the Christmas Bird Count in Kingsville, Texas Utilizing GIS Techniques

• The National Audubon Society's Christmas Bird Count
• Great Reef Count (fish)
• Fourth of July Butterfly Count (butterflies)

See Audubon’s CBC bibliography and the Approaches to analyzing your data section above

Anonymous. 1916. Bird-Lore’s Seventeenth Christmas Bird Census. Bird Lore 18:363-364.

Anonymous. 1907. Bird-Lore’s Seventh Christmas Bird Census. Bird Lore 9:16-31.

Anonymous. 1900. A Christmas Bird-Census. Bird Lore 2:192.