Effects of ignoring survey design information for data reuse

Abstract

Data are currently being used, and reused, in ecological research at an unprecedented rate. To ensure appropriate reuse however, we need to ask the question: "Are aggregated databases currently providing the right information to enable effective and unbiased reuse?" We investigate this question, with a focus on designs that purposefully favour the selection of sampling locations (upweighting the probability of selection of some locations). These designs are common and examples are those designs that have uneven inclusion probabilities or are stratified. We perform a simulation experiment by creating datasets with progressively more uneven inclusion probabilities, and examine the resulting estimates of the average number of individuals per unit area (density). The effect of ignoring the survey design can be profound, with biases of up to 250% in density estimates when naive analytical methods are used. This density estimation bias is not reduced by adding more data. Fortunately, the estimation bias can be mitigated by using an appropriate estimator or an appropriate model that incorporates the design information. These are only available however, when essential information about the survey design is available: the sample location selection process (e.g. inclusion probabilities), and/or covariates used in their specification. The results suggest that such information must be stored and served with the data to support meaningful inference and data reuse.

Climate change reshuffles northern species within their niches

“…S4), suggesting no changes in the spatial aggregation of sites over time. Furthermore, our analyses are model based and thus explicitly account for the number and distribution of sampling sites, while making inference on both environmental covariates and spatial random effects 50 . Thus, the variation in the number and distribution of sampling sites affects the uncertainty on trend estimates, rather than affecting the estimates themselves.…”

Joint species distribution modeling with competition for space

“…Notably, the spatially homogeneous models C+BB and C+DM did not differ significantly from the other models in per‐species total over area predictions ($$$ C{V}_3 $$$ in Table 3 and Figure B9), which is reasonable since, in theory, sample mean over uniform random locations is an unbiased estimate for species‐wise total covers. However, Foster et al (2021) give an example on how sample mean over uniform random locations may fail in practice if the underlying field is very heterogeneous. We expect that estimating areal vegetation composition more precisely can be beneficial in areal greenhouse gas emission models since vegetation is important explanatory factor in carbon dioxide and methane flux models.…”

Spatial distribution of biodiversity citizen science in a natural area depends on area accessibility and differs from other recreational area use

“…Further, it may be possible to incorporate local‐scale sampling bias within citizen science hotspots into regional models. A better understanding of sampling process can support a diverse range of applications that are relevant to local area management, including biodiversity assessments, monitoring of trends, assessment of interventions, invasive species detection, and species distributions analyses (Dobson et al., 2020; Foster et al., 2021; Johnston et al., 2022; Kühl et al., 2020).…”