Plan a Bird Count, Avoid Duplicate Tallies and Summarize Results

A useful bird count starts before you raise your binoculars. Decide where the session begins, how long it will last, whether you will stay put or move, and who is counting. Record numbers as you observe, and keep a running total for each species. Count conservatively, so a bird that flies off and returns is not tallied again. At the end, summarize each session on its own, with its date, location, duration and method attached. That approach follows eBird's counting guidance, and it works whether you submit to eBird, run a youth or classroom project, or keep a personal log of visits to the prairie and wetlands around Medicine Lake in northeastern Montana.

Define the session before you count

Under eBird guidance, one checklist covers one calendar date, one map location and one protocol, meaning the counting method that matches what you actually did. Treat each session the same way, even if you never upload it. A session that covers two dates, or that mixes standing still with walking, is hard to interpret later.

Start a new session when the location, the major habitat or your activity changes. On mixed prairie and wetland ground, a walk from upland grassland down to a marsh edge or open-water shoreline is a habitat change. Splitting the count at that point keeps grassland birds and wetland birds tied to the places they were found.

If you bird as a group, keep one shared list. If the group splits into separate parties, each party keeps its own list. Do not merge their observations afterward, because you can no longer tell which birds were seen where, or whether both parties counted the same individuals.

Choose a counting protocol

The protocol describes the effort behind the numbers. eBird uses the following categories, and they are a sensible framework for any count.

Protocol	When it applies	Effort details to record
Stationary	Birding is your primary purpose and you stay within about 30 meters of the starting point	Start time and duration
Traveling	Birding is your primary purpose and you move beyond that radius	Start time, duration and distance covered. If you backtrack, report only the unique distance of the route, not the total distance walked.
Incidental	You notice birds while doing something else, such as driving or working	No full effort details; the list is not a complete survey
Historical	You lack the required effort details, for example when entering old notes	Whatever details are known

For a count you plan to compare across visits, use Stationary or Traveling and aim for a complete list. A complete list includes every species you confidently identified by sight or sound. It does not count as complete if you left out species you found uninteresting or forgot to note.

Set up a worksheet

A tally sheet works best with two parts. A header holds the details that apply to the whole session. Detail rows hold each observation and leave room to flag possible repeats. Here is an original layout you can adapt on paper or in a spreadsheet:

SESSION HEADER
Session ID:            Date:
Location or route:
Start time:            End time:
Observers:
Protocol:              (Stationary / Traveling / Incidental / Historical)
Unique distance:       (Traveling only)
Conditions:

DETAIL ROWS
Time | Bird label | Count | Status (exact / estimated / X) | Counting increment | ID uncertainty | Movement | Possible duplicate of row #

The last three columns do the most work. Note which direction a bird or flock was moving, and link a row to an earlier one if it might be the same birds. Settle those links before you add to the species total. If a link cannot be resolved, keep the conservative count and leave the note in place.

Record counts in the field

• Write counts down during the session rather than reconstructing them from memory afterward.
• Keep a running total for each species and update it as new birds appear.
• Record an exact count when you can. When you can't, record an estimate and mark it as one.
• Use X for a species you identified but could not count. An X still makes the list complete, but it adds nothing to numerical totals.
• Assess each identification critically. Mark uncertain birds as uncertain instead of assigning them to a species.

Avoid counting the same bird twice

Duplicate tallies are the most common way counts get inflated. The rule of thumb in eBird guidance is to stay conservative. If a bird that just flew off is likely the one that has returned, do not count it again. Count it as a new bird only when something shows it is a different individual, such as distinctive markings or a different age or sex.

• Birds that come and go from one spot: at a feeder or any small gathering point, record the highest number visible at the same moment. Then add only individuals that are clearly different from those already counted.
• Birds that move along with you: on a traveling count, note the direction of birds that fly ahead or circle back. Link them to earlier rows instead of logging them as new.
• Shared lists: on a group count, call out sightings so one person can record them. Several people counting the same birds independently can multiply the total.
• Unresolved cases: if you cannot tell whether a sighting is a repeat, leave the total unchanged and record your reasoning in the notes.

Estimate flocks honestly

Large flocks are common on open water and in fields. Rather than guessing a total, count a block of birds, such as 10 or 50. Judge how many blocks of that size fill the flock, then multiply. Be careful when the flock is deep or unevenly packed. Birds hidden behind others, or a dense center with thin edges, make a simple block multiplication unreliable. Adjust your estimate for those areas, or count them as separate blocks.

For birds passing in a steady stream, you can estimate a rate and extrapolate. This only works while the flow stays consistent. When the stream speeds up, thins out or stops, make a new estimate for that period. Do not stretch the first rate across the whole session.

Keep the precision honest. If you estimated in blocks of 10, report a round figure such as 120. Do not add 3 stragglers to make it 123, because that implies an exact count you never made. Record the counting increment in your worksheet so the precision is clear later.

Summarize each session

Summaries should stay tied to the session that produced them. Total the accepted counts for each species after resolving duplicates. Carry the session's date, location, duration and protocol into the summary itself. That information should travel with the numbers, not sit in a separate file.

Hypothetical example: a 20-minute stationary session produces accepted tallies of 4 for Species A, 3 for Species B and 2 for Species C, for a total of 9 birds. A bar graph of those three tallies should include a caption stating the 20-minute duration and the stationary method. A reader can then see what effort produced the bars.

If you plan repeat visits, choose a consistent location, duration and method in advance and record them. When a visit departs from the plan, record what actually happened. Suppose one session runs 40 minutes instead of 20. Label it as a 40-minute count rather than placing it beside the 20-minute sessions as if the effort were equal. A longer or wider search usually finds more birds, so differences in effort must stay visible in any comparison.

Read the totals for what they are: the birds detected during that effort, at that place and time. A session total does not measure how many birds live in an area. A higher total on a later visit does not by itself show that a population has grown.

Count responsibly

Get access permission wherever it is required before you count. Keep enough distance that birds are not disturbed, especially near nests and roosts. If your approach flushes birds, you have both disturbed them and increased the risk of counting the same individuals twice as they resettle.