hurdle model (Poaceae)

Knowledge graph centered on hurdle model (Poaceae) with 53 nodes and 153 connections. Top connected: Rocky Mountain Biological Laboratory, Gothic, Two-form Bumble Bee, Bumble Bees, Yellow-fronted Bumble Bee.

Method synopsis

Two-part statistical model using Bayesian methods to analyze both presence/absence and amount of damage, accounting for zero-inflation in damage data.

Synthesized from method descriptions across 1 paper using this protocol.

Procedure from a recent peer-reviewed implementation

Steps below were extracted from the most recent peer-reviewed implementation of this protocol in the corpus — Herbivory damage but not plant disease under experimental warming is dependent on weather for three subalpine grass species (2023), Journal of Animal Ecology. Implementations in other papers (listed below) may differ.

1. Statistical modeling
   Hurdle-model approach was used in which presence or absence of damage was first modeled as a Bernoulli process, then conditional on the presence of damage, the amount of damage was modeled using zero-truncated observation. Damage was averaged across tillers to the individual plant level. Amount of damage (logit transformed) was modelled with a normal error distribution. Each grass species was analysed separately. Bayesian estimation was used in JAGS with the R2jags package. Models were run for at least 30,000 iterations with three chains, thinned by 5, and the first 15,000 iterations used as burn-in. Two separate equations were fit: (1) main effects model with warming treatment, days within year, and year effects; (2) interaction model adding fine-scale weather variables (temperature, 2- or 4-week precipitation) and their interactions with warming treatment.
   Quantities: 30,000+ MCMC iterations, 3 chains, thinning by 5, 15,000 burn-in iterationsDuration: Statistical analysis phaseConditions: Computational analysis
   Equipment: JAGS software, R2jags package, R statistical software

Typical Equipment

• JAGS software
• R2jags package

Output Measurements

• damage presence probability
• damage amount estimates
• treatment effects

Papers Using This Protocol (1)

Species Studied With This Protocol

Places Where Used