← Atlas catalogue Version V2 Version V1 Sources and limits V3 · Sonnet

Zastawie · Bagno Bubnów · a model of an evening, not an observation

The cranes' autumn assembly

Cranes return for the night in small skeins as the light fades. The observer counts for as long as they can see, and everything after that moment is silence in the record, not proof of emptiness.

Three invented evenings on a single sunset axis

Three horizontal time axes, from 60 minutes before sunset to 40 minutes after. On each axis, bars for the crane groups hang from the horizon line; a vertical amber line shows the end of the count, and the striped area beyond it marks time that was not counted. The full data are in the table Register of three evenings.

A teaching model, not data from the tower at Zastawie. Axis: minutes relative to sunset (T 0) on that evening. A bar is one recorded group, and its length is the number of birds (the same scale on all three axes).

Without JavaScript the axis and slider are disabled. The full record of all groups is in the register below, and the totals above correspond to an end of count at T +20 min.

A · Clear dusk sunset c. 18:45

lower 64upper 64

B · Fog and a possible repeat sunset c. 18:30

lower 58upper 92

outside the total: 1 entry after the end of count

C · Empty floodplain wetland sunset c. 18:15

lower 3upper 3

End of count

T +20 minafter sunset

Move the slider (arrow keys, Home, End) or choose a moment with a button. Entries up to and including that minute go into the totals; later ones stay in the record but outside the tally.

End of count T +20 min: A 64, B 58–92, C 3 (lower–upper, birds in the register).

Selected group

Select a bar on the axis (or the ‘Show on axis’ button in the register) to see the record and the limits of identification.

Register of three evenings

The same as on the axis, as a plain list with times relative to sunset. All the evenings are invented for the purposes of the model. The ‘In the tally’ column changes with the end of count and with the repeat markings at the top of the page.

Recorded groups: time relative to sunset, sector of the diagram (A–D from the left to the right of the field of view, with no map value), number of birds and status in the totals. Default state: end of count at T +20 min.
EntryTimeSectorRecordBehaviour and limitsIn the tally
A · Clear dusk · sunset c. 18:45 (model)
A1 18:10 T −35 min A 12 birds A regular skein, descending towards the water.High contrast before sunset. In both totals
A2 18:33 T −12 min B 28 birds A wide, spread-out line, loud calling before landing.Landing in the centre of the floodplain wetland. In both totals
A3 18:47 T +2 min C 9 birds A quiet, low flight over the scrub.Flight just after sunset. In both totals
A4 18:59 T +14 min A 15 birds A group visible in the afterglow of dusk.Recorded before the conventional end of count at T +20 min; with an earlier end it drops out of the totals. In both totals
B · Fog and a possible repeat · sunset c. 18:30 (model)
B1 17:50 T −40 min A 18 birds A family group feeding at the edge.An early flight in good light. In both totals
B2 18:24 T −6 min B 30–34 birds A flock approaching to land in the centre of the floodplain wetland.First record of the flock in sector B. The range 30–34 means the observer did not settle on an exact number. In both totals
B3 18:38 T +8 min B 30 birds A flock, flushed, circles and settles in the same sector.The observer suspects it is the same group as B2 (an assumption of the model, not a conclusion). The record stays; only its share in the totals changes. Upper total only (possible repeat)
B4 18:46 T +16 min D 10 birds A small group landing in the half-dark.Silhouettes visible against the light. In both totals
B5 18:56 T +26 min C voice only, no number Calling from the fog after dusk.Voice contact only; the number cannot be determined. This is not zero birds. Outside the tally – voice contact after the end of count
C · Empty floodplain wetland · sunset c. 18:15 (model)
C1 17:55 T −20 min A 3 birds A single family passing through without landing.A transit flight; they did not settle on the floodplain wetland. The model gives no reason why nothing else arrived. In both totals

How to read the totals

The two numbers for each evening are not a result but two honest readings of the same record.

An entry is not a bird

The observer records groups. A flock flushed from the reeds and circling over the same sector may be recorded twice. The number of entries does not equal the number of individuals.

Lower and upper

Lower leaves out groups marked as a possible repeat. Upper counts each as separate and takes the upper end of the range. These are the bounds of the tally from the register, not a confidence interval and not the size of the flock at the roost.

Dusk closes the count, not the arrival

The end of count is set by the observer, not by an astronomical formula. Poorer visibility does not mean a smaller arrival, and a ‘voice only’ entry neither adds a single bird to the total nor is it a zero.

Example from evening B (default settings: end of count T +20 min, B3 marked as a repeat).

Lower 58 = B1 18 + B2 30 + B4 10.

Upper 92 = 58 + 4 (upper end of the B2 range) + 30 (B3). The difference of 34 is all the uncertainty of this evening, and entry B5 (voice only, T +26) lies after the end of count and enters neither total.

Unmark B3 on the axis: the lower total rises to 88, because the same group is counted twice. Move the end of count before T +16 and B4 drops out of both totals.

On the tower, not at the roost

Brief and without a quiz. The rules and time windows below were carried over from the previous version of the page, together with the references [1] [2]. to be checked in the source

Behaviour towards the birds

  • Stay on the tower and look through binoculars. If your presence disturbs the birds, increase the distance or end the observation.
  • Do not look for a way closer to the floodplain wetland. The OTOP code recommends keeping to the routes and paths that are open to the public, and at a distance that does not scare birds at roosts.
  • Do not play recorded calls from a phone. Luring can disturb the birds, and an arrival has to be waited out.

When to look at all

  • Second half of August: according to the Polesie National Park guide, the start of the described window for gatherings. It is not a promise of a large flock on any given evening.
  • September: a window from about 15 August to the end of September, a reference point for planning, not information about the current flock.
  • October: the park describes birds staying even until the end of October in a warm autumn. Do not treat this as a standing extension of the season.
  • Outside this period: do not expect an evening assembly; plan to get to know the trail and the landscape instead.

Before you travel, check the park’s notices, the weather and the rules for access to the area. This page does not give the location of the roost or a route to it.

Sources and limits of the evidence

  1. Poleski Park Narodowy: ścieżka przyrodnicza „Czahary”, przewodnik dydaktyczny [Polesie National Park: ‘Czahary’ nature trail, educational guide]. Reference from V2: location of the tower at Zastawie, the Bagno Bubnów floodplain wetland, the window of autumn gatherings, the evening rhythm of the assembly. content not checked in V3
  2. OTOP: Kodeks Osób Obserwujących Ptaki [OTOP: Code of Conduct for Bird Watchers]. Reference from V2: keeping a distance, not luring birds with sound, noting uncertainty reliably. content not checked in V3

What the data on this page are

Evenings A, B and C are demonstration models for learning to read such a record. They are not measurements from any day, and the sunset times and group numbers are illustrative, not a distribution or a statistic. Sectors A–D are a conventional division of the field of view, not a map. The axis and bars are drawn in HTML and CSS as a schematic.

What changed relative to V2

Removed: the form for a personal notebook saved in the browser (its role was taken over by marking repeats on the axis), the quiz about behaviour, the choice of seasonal window (replaced by a list) and a photograph of cranes from another country. Kept: the lower/upper tally, the exclusion of repeats and the rule that a voice without a number adds no birds. This page contains no sound.