Create your own conference schedule! Click here for full instructions

Abstract Detail


Ruffley, Megan [1], Jimenez, Ivan [2].

Effects of Sampling Effort on Estimates of the Mean of Range Size Distributions.

The range size distribution (RSD) for a given region describes the frequency of species exhibiting geographic ranges of different sizes. Descriptions of spatial variation in RSD are major elements of current attempts to explain spatial patterns of diversity and identify areas of conservation concern. Thus, it is key to understand potential bias in estimates of RSD properties. Here, we hypothesize that bias in estimates of the mean of RSD is determined by sampling effort according to a model of the probability of not discovering a species. Based on this model we predict that bias in estimates of the mean of RSD for any given region is negatively related to mean sampling effort across the region, and positively related to the spatial aggregation of sampling effort across the region, because in poorly sampled areas narrowly distributed species are less likely to be discovered than widely distributed species. We tested these predictions in the context of the current belief that mean RSD of Andean plants is smaller than that of Amazonian plants. Botanical sampling effort across Northwestern South America was measured as “collector days”, defined as unique combinations of collector name and collection date. Aggregation in sampling effort was measured as the log of variance in collector days across grid cells conditional on mean sampling effort. We simulated the geographic distributions of virtual species across the Neotropics using the “stepping stone” model described in, and the discovery of each of the virtual species according to observed sampling effort and equation 1 with detectability = 0.1. We calculated bias in the mean of RSD as the difference between true and observed values of the mean of RSD, the latter being the mean AOO of discovered species excluding the undiscovered. The results supported both predictions on bias in estimates of the mean of RSD with p-values < 10-5 and < 3 x 10-4, respectively. Thus, current descriptions of geographic variation in RSD and the density of narrowly distributed plant species across the Neotropics may be more fiction than substance, and should be regarded as highly tentative at best. In future studies, estimates of spatial variation in mean RSD must account for spatial variation in sampling effort.

Log in to add this item to your schedule

1 - Miami University, Biology, Oxford, OH, 45056, USA
2 - Missouri Botanical Garden, CCSD, P.O. Box 299, St. Louis, MO, 63166, USA

none specified

Presentation Type: Poster:Posters for Topics
Session: P
Location: Hall D/The Shaw Conference Centre
Date: Monday, July 27th, 2015
Time: 5:30 PM
Number: PBG006
Abstract ID:742
Candidate for Awards:None


Copyright 2000-2015, Botanical Society of America. All rights reserved