Can School Children Support Ecological Research? Lessons from the Oak Bodyguard Citizen Science Project

18.03.2020

Scientific knowledge in the field of ecology is increasingly enriched by data acquired by the general public participating in citizen science (CS) programs. Yet, doubts remain about the reliability of such data, in particular when acquired by schoolchildren. We built upon an ongoing CS program, Oak Bodyguards, to assess the ability of schoolchildren to accurately estimate the strength of biotic interactions in terrestrial ecosystems. We used standardized protocols to estimate attack rates on artificial caterpillars and insect herbivory on oak leaves. We compared estimates made by schoolchildren with estimates made by professional scientists who had been trained in predation and herbivory assessments (henceforth, trained scientists), and trained scientists' estimates with those made by professional scientists with or without expertise (untrained) in predation or herbivory assessment. Compared with trained scientists, both schoolchildren and untrained professional scientists overestimated attack rates, but assessments made by the latter were more consistent. Schoolchildren tended to overestimate insect herbivory, as did untrained professional scientists. Raw data acquired by schoolchildren participating in CS programs, therefore, require several quality checks by trained professional scientists before being used. However, such data are of no less value than data collected by untrained professional scientists. CS with schoolchildren can be a valuable tool for carrying out ecological research, provided that the data itself is acquired by professional scientists from material collected by citizens.

Full text

Castagneyrol, B., Valdés-Correcher, E., Bourdin, A., Barbaro, L., Bouriaud, O., Branco, M., ... & Tack, A. J. (2019). Can school children support ecological research? Lessons from the 'Oak bodyguard'citizen science project. bioRxiv, 712638.


Location of oak trees included in the study. An interactive version of this map can be found in the Supplemental File as Figure 1.


Precision and accuracy of school children (a) and professional scientists (b) in assessing attack rate (% artificial larvae with predation marks). Dots represent attack rate aggregated at the level of oak trees for each survey separately. Dot size is proportional to the number of overlapping dots. Dashed lines indicate a 1:1 relation. In Panel a, the thick dashed red line represents the non-significant regression line (y = 0.08·x + 50.32, marginal R2: Rm2 < 0.01, conditional R2: Rc2 = 0.66). In Panel b, the bold red line represents the significant regression line (y = 0.66·x + 23.41, Rm2 = 0.31, Rc2 = 0.78). EVC, Elena Valdés Correcher (a single professional scientist trained to identify predation marks on artificial larvae).


Comparisons between insect herbivory as estimated by school children, trained scientists, and untrained professional scientists. Empty dots represent individual observations (i.e., a single assessment on a particular leaf set). Filled circles and vertical bars represent means ± SE of the raw data.