275
The Opportunities for Counting Birds in the Future
The true potential of ImageJ will be realised when used in conjunction with images
collected by light aircraft or drones, particularly in relation to inaccessible, or difficult to see, feeding or breeding grounds (see Chapt. 15). Furthermore, thermal
imaging cameras mounted on drones will offer the option of carrying out low disturbance night surveys of ground-nesting bird colonies, again providing images
suitable for processing in ImageJ.
The manual counter in ImageJ is a deceptively powerful tool and there is no
doubt that it has the potential to revolutionise how we collect and process bird count
data.
Conclusions
Bird surveillance schemes are among the most heavily subscribed citizen science
projects in the UK. Several of these are long-running national schemes such as the
Bird Atlas 2007–2011 (Balmer et al. 2013a, b), the Wetland Birds Survey (e.g. Frost
et al. 2016) and the Seabird Monitoring Programme (e.g. JNCC 2015). The BTO
coordinates the Wetland Bird Survey (WeBS), which has been operational since
1947, while the Seabird Monitoring Survey is organised by JNCC and has been
operational since 1986.
All of these rely heavily on untrained volunteers to collect data on large numbers
of birds, including wildfowl, waders, seabirds and farmland birds, yet there is nothing in the literature to suggest that the data provided by the volunteers is subject to
any form of validation. There is also nothing to suggest that photographic counts
have contributed to the data collection process, despite the longevity of the schemes.
The results of the observer variation exercises in this chapter suggest not only
that some validation of ‘by eye’ counts is essential, but also that the ornithological
community would benefit from embracing new technologies to improve both the
accuracy and precision of bird counts in the future.
Acknowledgements Thanks are due to Dan Guest, Chris Millican and Jonas Grahn for helping
with the thankless task of copy-editing and for providing invaluable comments on the content. I am
also grateful to the 35 experienced ornithologists who participated in the observer variation exercise by contributing flock size estimates of the images presented in this chapter: as agreed, these
contributors will remain nameless. Finally I would like to thank Ricardo Diaz-Delgado, Manuel
Mañez and their colleagues at Estación Biológica de Doñana for giving me the opportunity to
accompany their wetland bird counters, both on the ground and in the air: an exercise that rekindled my interest in the subject area and inspired me to write this chapter.
Improving the Accuracy of Bird Counts Using Manual and Automated Counts…
The Opportunities for Counting Birds in the Future
The true potential of ImageJ will be realised when used in conjunction with images
collected by light aircraft or drones, particularly in relation to inaccessible, or difficult to see, feeding or breeding grounds (see Chapt. 15). Furthermore, thermal
imaging cameras mounted on drones will offer the option of carrying out low disturbance night surveys of ground-nesting bird colonies, again providing images
suitable for processing in ImageJ.
The manual counter in ImageJ is a deceptively powerful tool and there is no
doubt that it has the potential to revolutionise how we collect and process bird count
data.
Conclusions
Bird surveillance schemes are among the most heavily subscribed citizen science
projects in the UK. Several of these are long-running national schemes such as the
Bird Atlas 2007–2011 (Balmer et al. 2013a, b), the Wetland Birds Survey (e.g. Frost
et al. 2016) and the Seabird Monitoring Programme (e.g. JNCC 2015). The BTO
coordinates the Wetland Bird Survey (WeBS), which has been operational since
1947, while the Seabird Monitoring Survey is organised by JNCC and has been
operational since 1986.
All of these rely heavily on untrained volunteers to collect data on large numbers
of birds, including wildfowl, waders, seabirds and farmland birds, yet there is nothing in the literature to suggest that the data provided by the volunteers is subject to
any form of validation. There is also nothing to suggest that photographic counts
have contributed to the data collection process, despite the longevity of the schemes.
The results of the observer variation exercises in this chapter suggest not only
that some validation of ‘by eye’ counts is essential, but also that the ornithological
community would benefit from embracing new technologies to improve both the
accuracy and precision of bird counts in the future.
Acknowledgements Thanks are due to Dan Guest, Chris Millican and Jonas Grahn for helping
with the thankless task of copy-editing and for providing invaluable comments on the content. I am
also grateful to the 35 experienced ornithologists who participated in the observer variation exercise by contributing flock size estimates of the images presented in this chapter: as agreed, these
contributors will remain nameless. Finally I would like to thank Ricardo Diaz-Delgado, Manuel
Mañez and their colleagues at Estación Biológica de Doñana for giving me the opportunity to
accompany their wetland bird counters, both on the ground and in the air: an exercise that rekindled my interest in the subject area and inspired me to write this chapter.
Improving the Accuracy of Bird Counts Using Manual and Automated Counts…