# Building a Reliable Citizen-Science Bird Survey

## Survey Design

Citizen-science bird surveys can produce valuable population data when volunteers follow a consistent protocol. In our spring survey, participants selected fixed observation points, recorded birds for exactly ten minutes, and submitted one checklist per visit. Repeating each point three times between April and June helped distinguish persistent patterns from weather-related variation.

## Data Collection

Observers used a mobile form to record species, counts, start time, coordinates, and detection method. The form also captured survey effort, including duration, distance traveled, and number of observers. These fields are essential: a checklist reporting five warblers during ten stationary minutes is not directly comparable to one collected during a two-hour hike.

```yaml
survey:
  duration_minutes: 10
  protocol: stationary
  repeat_visits: 3
  coordinate_precision_m: 25
```

## Quality Control

Automated validation flagged records outside known seasonal ranges, unusually high counts, coordinates beyond the study boundary, and incomplete effort data. A flag did not mean a sighting was wrong; it prompted a reviewer to request photographs, audio, field notes, or confirmation from another observer. This approach preserved credible unusual observations without allowing simple entry errors to distort the dataset.

## Analysis

We summarized detections by site and survey round, then modeled species occurrence while accounting for observation effort, time of day, and recent rainfall. Because non-detection does not prove absence, repeated visits were especially important. Sites surveyed only once were retained for descriptive maps but excluded from occupancy estimates requiring replication.

## Lessons Learned

Clear instructions and rapid feedback improved data quality more than complex statistical corrections. Short training videos, example checklists, and reminders about reporting zero detections reduced ambiguity in the final dataset. Publishing the protocol, validation rules, and anonymized observations also made the project easier to audit, reproduce, and extend in future seasons.