Accounting for imperfect detection in Hill numbers for biodiversity studies

Broms, Kristin; Hooten, Mevin B.; Fitzpatrick, Ryan M.

doi:10.1111/2041-210x.12296

Cited by 38 publications

(41 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…While eDNA representativeness assessment might be too complex to model (Alberdi et al, ; Barnes & Turner, ), amplification biases can be measured in silico (Piñol et al, ) and using mock communities (Lamb et al, ). This enables implementing occupancy‐modelling approaches that account for the relative amplification probability of OTUs (Ficetola et al, ), which have also been implemented within the framework of Hill numbers (Broms, Hooten, & Fitzpatrick, ; Iknayan, Tingley, Furnas, & Beissinger, ). These approaches are still in its infancy, but are likely to undergo a rapid development in the upcoming years.…”

Section: Dealing With Zero‐inflated Insufficient and Biased Datamentioning

confidence: 99%

A guide to the application of Hill numbers to DNA‐based diversity analyses

Alberdi

Gilbert

2019

Molecular Ecology Resources

174

138

View full text Add to dashboard Cite

With the advent of DNA sequencing‐based techniques, the way we detect and measure biodiversity is undergoing a radical shift. There is also an increasing awareness of the need to employ intuitively meaningful diversity measures based on unified statistical frameworks, so that different results can be easily interpreted and compared. This article aimed to serve as a guide to implementing biodiversity assessment using the general statistical framework developed around Hill numbers into the analysis of systems characterized using DNA sequencing‐based techniques (e.g., diet, microbiomes and ecosystem biodiversity). Specifically, we discuss (a) the DNA‐based approaches for defining the types upon which diversity is measured, (b) how to weight the importance of each type, (c) the differences between abundance‐based versus incidence‐based approaches, (d) the implementation of phylogenetic information into diversity measurement, (e) hierarchical diversity partitioning, (f) dissimilarity and overlap measurement and (g) how to deal with zero‐inflated, insufficient and biased data. All steps are reproduced with real data to also provide step‐by‐step bash and R scripts to enable straightforward implementation of the explained procedures.

show abstract

Section: Dealing With Zero‐inflated Insufficient and Biased Datamentioning

confidence: 99%

A guide to the application of Hill numbers to DNA‐based diversity analyses

Alberdi

Gilbert

2019

Molecular Ecology Resources

174

138

View full text Add to dashboard Cite

show abstract

“…Assuming no false positives, only zeros can be recorded for a species at sites where it is absent, hence the multiplication by z ik above, in the probability of the Bernoulli. The estimated presence/absence indicators (latent variables z 's) and occupancy probabilities ( ψ ’s) for individual species can be used to derive estimates of diversity metrics (Broms et al, ; Dorazio et al, ; Dorazio, Royle, Söderström, & Glimskär, ). For instance, site‐specific predictions of richness can be obtained by computing the expected number of species at each site i , that is, the sum of estimated occupancy probabilities:

{\overset{N}{true}}_{i} = \sum_{k} {\overset{ψ}{true}}_{ik}

.…”

Section: Modeling Approach and Interpretation Of Estimatesmentioning

confidence: 99%

Inferring species richness using multispecies occupancy modeling: Estimation performance and interpretation

Guillera‐Arroita

Kéry

Lahoz‐Monfort

2019

Ecology and Evolution

View full text Add to dashboard Cite

Multispecies occupancy models can estimate species richness from spatially replicated multispecies detection/non‐detection survey data, while accounting for imperfect detection. A model extension using data augmentation allows inferring the total number of species in the community, including those completely missed by sampling (i.e., not detected in any survey, at any site). Here we investigate the robustness of these estimates. We review key model assumptions and test performance via simulations, under a range of scenarios of species characteristics and sampling regimes, exploring sensitivity to the Bayesian priors used for model fitting. We run tests when assumptions are perfectly met and when violated. We apply the model to a real dataset and contrast estimates obtained with and without predictors, and for different subsets of data. We find that, even with model assumptions perfectly met, estimation of the total number of species can be poor in scenarios where many species are missed (>15%–20%) and that commonly used priors can accentuate overestimation. Our tests show that estimation can often be robust to violations of assumptions about the statistical distributions describing variation of occupancy and detectability among species, but lower‐tail deviations can result in large biases. We obtain substantially different estimates from alternative analyses of our real dataset, with results suggesting that missing relevant predictors in the model can result in richness underestimation. In summary, estimates of total richness are sensitive to model structure and often uncertain. Appropriate selection of priors, testing of assumptions, and model refinement are all important to enhance estimator performance. Yet, these do not guarantee accurate estimation, particularly when many species remain undetected. While statistical models can provide useful insights, expectations about accuracy in this challenging prediction task should be realistic. Where knowledge about species numbers is considered truly critical for management or policy, survey effort should ideally be such that the chances of missing species altogether are low.

show abstract

“…). We also showed how to derive commonly used community metrics through Hill numbers (Broms, Hooten & Fitzpatrick ), stratifying them by functional groups. Deriving biodiversity metrics while accounting for imperfect detectability is possible only in a multi‐species occupancy framework, and cannot be accommodated in traditional approaches like rarefaction and extrapolation procedures (Chao et al .…”

Section: Discussionmentioning

confidence: 99%

“…In addition, community‐ and guild‐specific diversity metrics can be easily computed. For example, using occupancy probabilities,

ψ_{i r}

, it is possible to directly estimate guild‐ and region‐specific Hill numbers (

^{q} Δ

; Broms, Hooten & Fitzpatrick ) that summarize three types of biodiversity metrics commonly used, (i) species richness ( q = 0), (ii) Shannon diversity ( q = 1, the Shannon entropy exponentiated), and (iii) Simpson diversity ( q = 2, inverse of the complement of the Gini–Simpson index; Jost ):

^{q} {normalΔ}_{g r} = {()(, \sum_{i = 1}^{N_{g r}}, {()(, \frac{ψ_{i r}}{\sum_{s = 1}^{N_{g r}} {normalψ}_{s r}})}^{q})}^{\frac{1}{1 - q}} q \neq 1, q \geq 0

and

^{1} {normalΔ}_{g r} = exp (- \sum_{i = 1}^{N_{g r}} \frac{{normalψ}_{i r}}{\sum_{s = 1}^{N_{g r}} ψ_{s r}} log \frac{{normalψ}_{i r}}{\sum_{s = 1}^{N_{g r}} ψ_{s r}}) q = 1 .

We demonstrate the derivation of these using our case study below, and provide the script for their derivation in the supplemental model code.…”

Section: Methodsmentioning

confidence: 99%

Quantifying spatial variation in the size and structure of ecologically stratified communities

et al. 2017

View full text Add to dashboard Cite

Summary Including ecological specialization (e.g. functional guild) in analyses performed across regions can help to study how size and structure of communities vary across environmental gradients. Multi‐species occupancy models, and their extension to a multi‐region framework, represent useful tools for such gradient analysis based on functional traits. However, in these models species richness is only a derived parameter and therefore explicit relationships cannot be inferred. We provide a novel hierarchical multi‐region community model that allows for direct modelling of trait‐based patterns of species richness along environmental gradients by partitioning communities into ecologically relevant strata (e.g. guilds). We illustrate the flexibility of the model by simulations, where we estimate guild richness and community composition using both categorical and continuous species traits. In addition, we apply our model to data from eight avian communities to explore how guild richness varies across elevational gradients. Simulations indicate that the model is able to estimate relationships between guild richness and environmental gradients even when the number of regions is low. Applying the model to the bird community data set, we show how the number of species in four feeding guilds varies in different ways across an elevational gradient. Understanding large scale variation in species richness and how it relates to environmental gradients requires an explicit treatment of community structure. Our modelling approach allows the testing of hypotheses directly and simultaneously on strata‐ and community‐specific species richness while accounting for the issues of rarity and detectability, and can be used to investigate the mechanisms which shape large scale spatial variation in species richness.

show abstract

Accounting for imperfect detection in Hill numbers for biodiversity studies

Cited by 38 publications

References 31 publications

A guide to the application of Hill numbers to DNA‐based diversity analyses

A guide to the application of Hill numbers to DNA‐based diversity analyses

Inferring species richness using multispecies occupancy modeling: Estimation performance and interpretation

Quantifying spatial variation in the size and structure of ecologically stratified communities

Contact Info

Product

Resources

About