Compound Poisson‐gamma vs. delta‐gamma to handle zero‐inflated continuous data under a variable sampling volume

Abstract

Summary1. Ecological data such as biomasses often present a high proportion of zeros with possible skewed positive values. The Delta-Gamma (DG) approach, which models separately the presence-absence and the positive biomass, is commonly used in ecology. A less commonly known alternative is the compound Poisson-gamma (CPG) approach, which essentially mimics the process of capturing clusters of biomass during a sampling event.2. Regardless of the approach, the effort involved in obtaining a sample (henceforth called the sampling volume, but could also include swept areas, sampling durations, etc.), which can potentially be quite variable between samples, needs to be taken into account when modelling the resulting sample biomass. This is achieved empirically for the DG approach (using a generalized linear model with sampling volume as a covariate), and theoretically for the CPG approach (by scaling a parameter of the model). In this study, the consequences of this disparity between approaches were explored first using theoretical arguments, then using simulations and finally by applying the approaches to catch data from a commercial groundfish trawl fishery. 3. The simulation study results point out that the DG approach can lead to poor estimates when far from standard idealized sampling assumptions. On the contrary, the CPG approach is much more robust to variable sampling conditions, confirming theoretical predictions. These results were confirmed by the case study for which model performances were weaker for the DG. 4. Given the results, care must be taken when choosing an approach for dealing with zero-inflated continuous data. The DG approach, which is easily implemented using standard statistical softwares, works well when the sampling volume variability is small. However, better results were obtained with the CPG model when dealing with variable sampling volumes.

Scaling in the Local Strain-Rate Field during Jerky Flow in an Al-3%Mg Alloy

“…ε a as belonging to a certain class of complex dynamical systems. Namely, the exponent 1 < p < 2 corresponds to compound Poisson-gamma distributions, which are particularly used to mimic the process of capturing clusters in ecological data such as biomasses [73]. Therewith, the observation of a crossover in the scaling for some deformation curves (Figure 7b) may be indicative of multifractality [74], which is in consistence with multifractal behaviors detected for the PLC effect [17,21,28].…”

Developing and assessing a density surface model in a Bayesian hierarchical framework with a focus on uncertainty: insights from simulations and an application to fin whales (Balaenoptera physalus)

“…Finally, we implemented a Tweedie distribution as a flexible model for count data that allows for both a large number of zeros and overdispersion, two features ubiquitous in animal population surveys. Because the Tweedie is not a built-in distribution in most Bayesian software packages, we adopted the CPG approach of Lecomte et al (2013). Together these features provide more options for users when fitting Bayesian models to line transect data.…”

The effect of the lunar phase on the catch per unit effort (CPUE) of the Turkish swordfish longline fishery in the eastern Mediterranean Sea

“…Given the existence of few zero catches, a Tweedie GAM following a compound Poisson-gamma approach was applied. This avoids multiple-stage modelling of zeroinflated data and allows to model jointly the probability of presence and the non-zero sampled quantity (Shono 2008, Lecomte et al 2013. The smoother function used was a penalized cubic regression spline and model fitting was accomplished using the "mgcv" library (Wood 2006) under the R language environment (R Core Team 2017).…”