Table 6 Comparison of the different generalized linear mixed models tested for analysing factors affecting the prevalence of co-infection in 641 P. plantaginis populations in Åland.

From: Co-infection alters population dynamics of infectious disease

Model

DIC

Range posterior estimate (posterior standard error)

Logistic regression

1,620

—

Spatial intercept

1,513

1,076 (±523)

Spatial logistic regression factor: age of the population

1,508

1,020 (±494)

Spatial logistic regression factor: pathogen-based connectivity

1,506

981 (±478)

Spatial logistic regression factor: number of multilocus genotypes

1,498

1,309 (±566)

Spatial logistic regression factor: disease abundance

1,492

1,253 (±570)

Spatial logistic regression factors: all

1,469

1,268 (±585)

  1. DIC, deviance information criterion.
  2. DIC used to determine the best model. Range represents the distance where the spatial autocorrelation drops below 0.1, according to the Matérn covariance function. The posterior estimates for the models including a spatial random effect are indicated.
Back to article page