Field investigations of phenotypic variation in free‐living organisms are often limited in scope owing to time and funding constraints. By collaborating with online communities of amateur naturalists, investigators can greatly increase the amount and diversity of phenotypic data in their analyses while simultaneously engaging with a public audience. Here, we present a method for quantifying phenotypes of individual organisms in citizen scientists’ photographs. We then show that our protocol for measuring wing phenotypes from photographs yields accurate measurements in two species of Calopterygid damselflies. Next, we show that, while most observations of our target species were made by members of the large and established community of amateur naturalists at iNaturalist.org, our efforts to increase recruitment through various outreach initiatives were successful. Finally, we present results from two case studies: (1) an analysis of wing pigmentation in male smoky rubyspots (Hetaerina titia) showing previously undocumented geographical variation in a seasonal polyphenism, and (2) an analysis of variation in the relative size of the wing spots of male banded demoiselles (Calopteryx splendens) in Great Britain questioning previously documented evidence for character displacement. Our results demonstrate that our protocol can be used to create high quality phenotypic datasets using citizen scientists’ photographs, and, when combined with metadata (e.g., date and location), can greatly broaden the scope of studies of geographical and temporal variation in phenotypes. Our analyses of the recruitment and engagement process also demonstrate that collaborating with an online community of amateur naturalists can be a powerful way to conduct hypothesis‐driven research aiming to elucidate the processes that impact trait evolution at landscape scales.
Dataset1
Dataset1
Contains data for comparing measurements made from perched individuals to those made from photographs taken under standardised conditions.
perched.hw = the total proportion of the isolated hindwing that is visible in perched photos
perched.tw = the total proportion of the surface of wings that is visible in perched photos
standard.hw = the total proportion of the hindwing surface that is pigmented from standardised photos
standard.tw = the total proportion of the surface of wings that is pigmented from standardised photos
Dataset2a
Contains data for observation accrual rate analyses
month.ordinal.upload = unique ID for month of upload
project.launch = whether observation comes from before or after the launch of the project
focal.species = 1 if the focal species (e.g. demoiselles)
month.upload = month of upload (raw)
observation.count = number of observations submitted during the month
project = iNaturalist project page corresponding to data
Dataset2B
Contains data for observation accrual rate analyses
month.ordinal.upload = unique ID for month of upload
project.launch = whether observation comes from before or after the launch of the project
focal.species = 1 if the focal species (e.g. demoiselles)
month.upload = month of upload (raw)
observation.count = number of observations submitted during the month
project = iNaturalist project page corresponding to data
new.recruit = 1 if individuals whose first observations in our dataset were of a focal species after the project launch date
Dataset2b.csv
Dataset3a
Contains data for case study
inaturalist.id = id of observation from iNaturalist.org
julian.date.wyear = julian date including year
julian.date = julian date used in analyses
month = month
latitude = decimal latitude
longitude = decimal longitude
site.id = unique ID for site
region = region assigned for analyses (for smoky rubyspot analyses, see main text)
scientific_name = scientific name
perched.tw = the total proportion of the surface of wings that is visible in perched photos
perched.hw = the total proportion of the isolated hindwing that is visible in perched photos
perched.hw_interpolated = perched.hw data with missing points interpolated from regression analyses
Dataset3b
Contains data for case study
inaturalist.id = id of observation from iNaturalist.org
julian.date.wyear = julian date including year
julian.date = julian date used in analyses
month = month
latitude = decimal latitude
longitude = decimal longitude
site.id = unique ID for site
sympatric = whether observation is from zone of sympatry for two demoiselles (for banded demoiselle analyses, see main text)
scientific_name = scientific name
perched.tw = the total proportion of the surface of wings that is visible in perched photos
perched.hw = the total proportion of the isolated hindwing that is visible in perched photos
perched.tw_interpolated = perched.tw data with missing points interpolated from regression analyses
perched.hw_interpolated = perched.hw data with missing points interpolated from regression analyses