Robust detection of plant species distribution shifts under biased sampling regimes

Abstract

Abstract. The great diversity of terrestrial plants testifies to a wide variety of life-history strategies for dealing with the problem of surviving to reproduce in a stressful environment while competing with other biota. It is essential to develop an understanding of how global environmental changes are perturbing the distribution of species throughout the globe. However, the most abundant collections of historic data on species distributions, i.e., museums and herbaria, are affected by sample biases that strongly compromise our ability to use these data for change detection studies. Here, we present a simulation study to find robust methods for rejecting spurious shifts in the geographic range or the environmental niche occupied by species, under a variety of sampling biases. We present two methods for addressing bias. The first method is a Bayesian weighting method from machine learning theory, in which each specimen is reweighted to achieve uniform sampling intensity in time, based on sampling intensity for all specimens. An alternative method uses binomial probabilities of selecting the observed number of samples of a target relative to all specimens, and estimates the likelihood associated with the binomial probabilities using Markov chain Monte Carlo (MCMC). We find that without dealing with sampling bias, using raw data is almost certain to provide inaccurate results, under even the mildest perturbations to idealized sampling. Among the two methods for addressing bias, the empirical estimate of the probability density of a change provided by MCMC was essential for accurately rejecting false changes. The performance of the weighting method, while an improvement over the use of raw data, was limited to cases in which the bias was weak. We conclude that species distribution changes are most robustly estimated using binomial probabilities, in which the probability space is explored empirically by MCMC.

Altitudinal shifts of the native and introduced flora of California in the context of 20th‐century warming

“…Testing for the presence of sample selection bias across an aggregated herbarium dataset is considered to be one of the most difficult issues confronted by researchers creating species range estimates (Elith et al ., ). To address this source of bias, we developed, validated (Wolf et al ., ), and employ here a robust statistical methodology to detect shifts in distributions over time. We also tested for bias in herbarium data introduced by collector behaviour—specifically collectors seeking ever‐higher‐elevation exemplars of particular taxa of interest (see http://onlinelibrary.wiley.com/doi/10.1111/geb.12423/suppinfo).…”

“…The estimation procedure follows Wolf et al . (), which describes in detail how to estimate the mean, variance, and estimated shift between sampling periods of a target species along an environmental gradient or geographical range. A concise overview of the approach is described in http://onlinelibrary.wiley.com/doi/10.1111/geb.12423/suppinfo.…”

Australian seaweed collections – Huisman and Millar respond

“…We (Huis man & Millar 2013) contended that this method could not be defended given the limited number of records, possible collector bias and lack of later collection effort from regions that had provided the earlier records. As noted by Wolf et al (2011), 'Using sample data without explicitly accounting for bias is almost guaranteed to provide strongly misleading results except under the most fortunate circumstances'.…”

“…apparently expanded), so the fewer collections in the later period cannot be explained by reduced abundance. To us, the most likely cause is sampling bias, which is a common feature of natural history collections (Wolf et al 2011) and one that is well known to us through our combined experience of over 70 years of algal collection and working in herbaria.…”

“…If herbarium records are to be used for modelling in this fashion, then perhaps it is time to halt this decline and aim to accumulate sufficient specimens to allow defensible assess ments of present range limits, and to subject historical records to analyses correcting for collection bias along the lines of those described by Wolf et al (2011).…”