CBS06 Line transects surveys of breeding birds throughout Konza Prairie


Like CPB01 and PBG051, this dataset includes records of bird species based on line transect sampling. These surveys re-initiate most of the original KNZ transects surveyed from 1981-2009 (initiated by Zimmerma), continue the two focal 300-m transects conducted from 2011-2016 by Sandercock on watersheds C3A, C3B, C3C, C1A, and 1D, (but not the peripheral segments which have been dropped), and adds two new 300-m transects to additional watersheds with the goals of (i) replicating all surveyed treatments at the watershed level, (ii) adequately sampling the grassland habitats present on Konza as of 2017, (iii) capturing transitions in bird community in response to woody encroachment and fire reversal treatments. The goals are to document and quantify long-term changes in the population sizes of breeding birds at Konza prairie by obtaining a standardized measure of annual breeding bird abundance at the site level and individual watershed or treatment level.

Data set ID: 


Core Areas: 

Short name: 



Original transects varied from ~250 m to ~1.5 km, typically bisected watersheds roughly in the direction of the longest axis. Although roughly linear, lines sometimes bent or wiggled. Original conduit posts are probably gone, but Brett Sandercock walked all transect lines and marked them with a GPS unit in 2011 or 2012. PBG and new transects were/are never marked physically on the ground. Apparently, no limits to the detections of birds that observers recorded were placed relative to distances to watershed boundaries or other transect lines. However, previously, observers made an attempt not count birds detected outside of watershed boundaries. When possible, new transects do not come closer than 100 m from a watershed boundary nor closer than 100 m perpendicular to other transects. When located ~end-to-end, transect lines are sometimes closer, but this should not pose a problem for double-counting birds (see below). New transects were located using GIS and aerial imagery to bisect geographic gradients within watershed and present a representative sample of the habitat within watersheds.

Each of the older CPB01 transects have been divided into 300 m units, and all new watersheds that are now surveyed under this protocol are exactly 300 m long. Where possible, there are at least 2, 300-m lines in each watershed that do not intersect one other and do bisect major topographic axes of the unit. These protocols try to preserve as much of those originally documented in the Konza data catalogue of 1980s as possible. Changes to the protocols reflect (a) updated equipment, (b) the desirability of obtaining spatial data for birds observed, and (c) lack of detail regarding environmental conditions under which summer bird surveys should be conducted. Regarding (c), I have incorporated rules of thumb communicated to me by Brett Sandercock as conducted on the patch-burn-grazing surveys in 2011-2016 and noted any changes to protocols that I am aware of (and the reason for those changes).

One observer counts birds on each transect once per year (typically Boyle, but occasionally, other skilled observers). They start counting birds ~30 min before sunrise and attempt to not begin a transect line after about 10:00 am. On a single morning, it is possible to survey 4-6, 300-m lines or roughly 2-3 km total (for irregular-length lines) in a single morning, providing that transects are not located too far away from one another. Previously, observers selected transects using a random procedure. Due to the large number of transects now being surveyed, logistical/time constraints now dictate which surveys are conducted in a single morning. Transects surveyed in a single morning minimize between-transect travel times, attempt to avoid walking into the sun, and do not survey birds on two watersheds of the same treatment on a single morning.

If possible, all transects are surveyed in the month of June. Originally, the handbook indicated that they should be run during the first 2 weeks of June. Actual survey dates range from 27-May to 28-Jun. Previous protocols indicated that observers should not start a transect when wind speed exceeded 10 mph (16 kph or no higher than Beaufort 3, see appendix 4). PBG051 did not have any explicit wind or temperature restrictions. From 2017 onward, surveys occur when wind conditions are up to and including Beaufort 4 but not higher. Also, because detection of many grassland birds is strongly affected by temperature, they are not initiated when the temperature is lower than 10*C or higher than ~26*C. Mist, drizzle, cloud cover is fine, but not fog or any precipitation more than a very light rain. Observers now note weather conditions at start and end of survey period using weather data from Manhattan.

Observers record information about survey events (observer, date, watershed, transect #, start and end time, weather at beginning and end) then a record of all birds detected. Observers note species using standard American Ornithological Society 4-letter codes, N individuals when birds are simultaneously detected in a group, not audio detection only, fly-overs (i.e. birds that do not touch down, forage over, or otherwise interact with the vegetation or ground), birds off-watershed, or beyond line, and sex of birds if known. Previously, the distance along the line was not noted, and only the perpendicular distance of the bird from the line was noted. To enable mapping of birds on the landscape, from 2017 on, observers record (a) their distance along the line, (b) the actual distance from the observer to the bird, (measured using a range-finder for at least 30% of birds, and estimated by eye when accuracy by comparing rangefinder and estimates are within ~5 m) and (c) the angle from the observer to the bird measured using a compass. Observers start by noting all birds detectable at the start of their line, then walk slowly, stopping as needed to record all birds detected as you go, pausing every 30 m or so to look and listen. The pace is was formerly noted as being ~1 km/45 min (PBG051) but due to recording angles and locations, a typical 300 m transect takes ~30 minutes.

The observer records the locations where they first detected the bird for any individual or group. For true flyovers, record the closest location to the transect line that the individual or group passed over. Where an aerial insectivore (e.g., nighthawk, swallow) is flying over the prairie low and likely foraging, do not count that as a flyover. If you detect an upland sandpiper way above a plot, try to determine where it is coming down. If it appears to be circling a territory on the watershed, count it NOT as a fly over. The point is to be able to exclude observations of birds not interacting with the local habitat with this data field. There is no limit on the distances at which you can record detections. It is very important NOT to double-count individual birds within or between transects! Some birds may be detected beyond the transect line (i.e. a perpendicular line drawn from the bird’s location to the angle of the line would bisect somewhere beyond the endpoint), but observers attempt to only record birds detected at unlimited distances in an imaginary rectangle extending to the L and R of each line. Previous protocols did not include birds beyond the line, so to facilitate comparisons to old data, only use the “beyond line” checkbox if you can’t help recording a rarer bird past the end of the line (backwards or forwards) or if you note a bird and then subsequently realize it is beyond bounds.

For additional metadata information see:
For additional methods information see:

Data sources: 




Subscribe to RSS - LTER