Geolocator wetness data accurately detect periods of migratory flight in two species of shorebird


112 – 119

1 August 17

Phil F. Battley, Jesse R. Conklin


Phil Battley
Ecology Group, Institute of Agriculture and Environment, Massey University, Private Bag 11-222, Palmerston North 4442, New Zealand


Members-Only Files

You must be an IWSG member to access these files. Already a member? Please log in.

While the principal use of light-recording geolocators is to determine geographical locations of migratory birds, supplementary wetness data have been used to refine estimates of minimum flight duration, on the assumption that a wet logger indicates the bird is on the ground. We provide a test of this assumption, by comparing wetness values against directly observed migratory departures of logger-equipped Bar-tailed Godwits Limosa lapponica and Red Knots Calidris canutus from the Manawatu River Estuary, New Zealand. Loggers recorded wetness every 10 min (Biotrack MK4093 and MK5093) or every one or four hours (Migrate Technology C65K). We retrieved loggers from 41 godwits from 2008–2014 and from seven Red Knots in 2013–2014 for which we had corresponding departure information; in total there were 51 departures of godwits and seven of knots that we could match to actual departure times (this included multiple years for some godwits). Overall, 10-min wetness data were very accurate for both godwits and knots (median estimated departure times were 14 min and one min later than true departure, respectively), as were the 60-min and 240-min loggers on godwits if corrected by the wet counts that are recorded within measurement intervals (medians of 16 min earlier and 8 min earlier, respectively). These longer-interval loggers were still reasonably accurate without this adjustment (medians of 37 min and 74 min later, respectively). There was substantial variation between individuals and logger types, with 10-min loggers going dry up to 148 min (godwit) or 55 min (knot) earlier than true departure, while the 60-min and 240-min loggers recorded wetness up to 142 min or 124 min later than true departure (or 195 min or 232 min later, if unadjusted). Some of this variation simply reflects the interval over which wetness is recorded, but bird behaviour and/or logger performance must play a role in some cases (e.g. the logger going dry before departure or remaining wet after departure). Given observed bird behaviour upon arrival after  migration (feeding on wet tidal flats), the wetness recording of geolocators is likely to give an accurate estimate of migratory flight duration, at least for species that frequent wet, particularly marine, habitats.