Structured and unstructured citizen science: Seven decades of expanding bird populations in central Ontario, Canada

Abstract

Citizens’ observations of wildlife offer a wealth of information. Such data, however, may be prone to biases or inaccuracies — in particular, the over-reporting of novel species and under-reporting of common or mundane species. Few studies have assessed the relationship between unstructured and structured observations using long-term data. We compared 69 years of voluntary bird sightings, recorded in the regular publication of Peterborough Field Naturalists (PFN), 1948–2016, to systematic Christmas Bird Counts (CBC; during winter) and Breeding Bird Surveys (BBS; during spring) in central Ontario, Canada. We focused on seven species that had expanded their ranges into the area: Northern Cardinal (Cardinalis cardinalis), House Finch (Haemorhous mexicanus), Purple Finch (Haemorhous purpureus), Red-bellied Woodpecker (Melanerpes carolinus), Wild Turkey (Meleagris gallopavo), Canada Goose (Branta canadensis), and American Robin (Turdus migratorius). For each species, we compared the first record, the annual number of sightings, and the annual relative abundance from the PFN to the first record and annual relative abundance in the CBC and BBS. We anticipated that naturalists’ accounts would be more sensitive to the arrival of novel species but would underestimate abundance as the species became common.In general, our results agreed with these predictions. In 11 of 12 cases, the naturalists detected a species well in advance of the CBC and BBS — on average by 1½ decades (range: 5–33 years). The correlations in bird abundance between unstructured (PFN) and structured (CBC and BBS) data were variable; most were weak and negative. We found significant negative correlations for the two most common species, Northern Cardinal (winter) and House Finch (spring), as well as for Wild Turkey (winter), and a positive correlation for the rarest species, the Red-bellied Woodpecker (spring). We surmise that novelty underlies these results — that rarity increases the likelihood of reporting by non-professionals. We conclude that unstructured observations are valuable in detecting novel species, but they may be inconsistent, even inverse, indicators of the abundance of expanding species, once established.