INVESTIGADORES
OJEDA valeria Susana
artículos
Título:
Extreme uncertainty and unquantifiable bias do not inform population sizes.
Autor/es:
ORIN JOHNSON +33 AUTORES; OJEDA, V.
Revista:
PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA
Editorial:
NATL ACAD SCIENCES
Referencias:
Lugar: Washington DC, USA; Año: 2022 vol. 119 p. 211386211 - 211386211
ISSN:
0027-8424
Resumen:
Species-specificpopulation estimates are fundamental for many aspects of ecology, evolution,and conservation, yet they are lacking for most species. Aiming to fill thisgap, Callaghan, Nakagawa, and Cornwell [1] estimated global bird population sizesby modeling the relationship between eBird reporting rates and independentestimates, and extrapolating globally. While we applaud their intention,we caution that their modeling framework is prone to yield extremely uncertainand biased estimates that cannot support robust inferences about speciesabundance distributions nor other applications in ecology, evolution, orconservation [1, 2]. Theirmethods yield extremely large posterior uncertainties fortotal global bird abundance (3.9B ? 2,080B; from [1] Figure 2), and 96% ofindividual species had posterior uncertainty spanning ≥3 orders of magnitude. GlaucousGull (Larus hyperboreus) was listed as the fifth commonest bird globally;it is difficult to be confident in this conclusion given that the 95% credibleinterval (CI) for Glaucous Gull overlapped the CIs for ~67% of all bird species.This uncertainty in species ordering makes it impossible to use these estimatesfor reliable conservation prioritization as suggested [1]. The tremendousuncertainty associated with the estimates of population size results from theinadequacy of the ten measures used to account for imperfect detection of birdsin eBird data [1], for which there is extreme inter- and intra-specificvariation in the observation process across regions, time, and habitat [6].eBird reporting rates also depend heavily on species? overlap with the activityof eBird users, which also varies by region, time, and habitat.In addition to highuncertainty, the approach also led to biased population estimates for manyspecies. Abundance estimates [1] fell outside minimum-maximum ranges providedby BirdLife International for 81% of the 2,423 species with available estimates(27% below the minimum, 54% above the maximum) [3]. Even the large uncertaintyintervals repeatedly failed to cover known true values. For example, SwiftParrot (Lathamus discolor) which was recently assessed at 280individuals [4], had a CI from 4,520 to 40M [1]. Spoon-billed Sandpiper (Calidrispygmaea) has a population size of 490 individuals (95% CI 360-620) [5], buthad a CI from 6,050 to 47M [1]. San Andres Vireo (Vireo caribaeus) usedin model training with a population size of 2,500-10,000 mature individuals [3],was estimated as extinct [1]. Regional differences inreporting rates create bias, because, as noted by Callaghan et al, the 7% of species usedto train the model were heavily biased towards Europe and North America [1]. Density imputation based on spatially uneven calibrationestimates biases the population estimates to an unknown and inestimable extent,with downstream influence on the shapes of species abundance distributions andecological conclusions.  For species withsufficient data quantity and quality, citizen and community science data canproduce reliable density estimates [6], and methods for such analyses areconstantly improving [7, 8]. However, no method currently exists to estimateglobal population sizes across species while accounting sufficiently for knownsources of variation in eBird reporting rates. Meaningful global populationestimates would represent a tremendous advance for ecology, evolutionarybiology, and conservation, but will require considerably more nuanced analysisof globally available data.