Seems Like a Random Place for a Bird Survey

By February 26, 2013Science

Before conducting field surveys, RMBO staff contact landowners for permission to access their land. Through these series of phone conversations and e-mail exchanges, a similar question from landowners often arises, “Why do you want to survey that particular location, especially when there are better birding spots nearby?”

A valid, however vague, answer would be, “It was picked at random.” But this statement does not tell the whole story. For that, you need a little statistics. I’m not a statistician, but as I learned in my first statistics class, the odds favor a large proportion of readers falling asleep if I get long-winded about the topic, so let me be brief.

Survey sites are picked randomly within a survey area to avoid bias. For example, imagine estimating the number of birds in a typical backyard. To compare this yard survey to the size and scope of many RMBO surveys, also imagine that you are the size of a flea and can only see a small portion of the yard at a time.

If you simply looked at the bird feeders in the yard, you will see many birds, and when applying your count of birds to the area of the entire yard, you would overestimate the number of birds as there are disproportionately more at the feeders. Conversely, if you only observe the undergrowth around one small tree, you may see few birds and underestimate the total population of the yard. But if you randomly observe several parts of the yard before constructing your estimate, and you have enough random samples, you will develop a clearer picture – thus the need for random survey locations.

How then, you might ask, does RMBO randomly select a site for survey? By tossing darts at a giant map of the Rocky Mountain region plastered on the side of our office? Sounds entertaining to me, but as you may have guessed, random site selection is performed digitally. With access to ever-improving digital satellite imagery and digital mapping programs such as ArcGIS and Google Earth, RMBO staff members employ computer algorithms to randomly select specific sites for surveys based on exact coordinates within the designated survey area.

Random Sampling Map

A map from an international restoration monitoring project shows how random survey points are chosen. Random points (numbered 1-12) are generated by a computer algorithm within the boundaries of the prescribed burn area (outlined in blue). The first four points that met project criteria (1-4, green points) were selected and 1-km transects were placed at random bearings from each (brown lines). Click map to enlarge.

If you are a landowner or partner who knows a good birding spot near one of our random survey sites, please do tell us about it. Field technicians may be able to briefly visit and record any rare species sightings, and as always will keep locations on private lands anonymous if desired. If you are interested in learning more about how our surveys work, please visit our page on bird population monitoring or contact a staff member at RMBO.

~ Bill Tiedje, Landowner Outreach and Program Technician