Canoco 5.12 is a minor update of Canoco 5.11. Canoco 5.12 primarily fixes some bugs. It also adds an estimation of significance levels to the calculation of species indicator values. See details.
Existing Canoco 5 users can update via Help|Check for Updates. For new orders check Updates and Orders.
Published on October 29, 2018
Canoco 5.11 is an update of Canoco 5.10, also known as Canoco 5.1. This update adds to Canoco contributions of response variables (e.g. species) to axes and contribution biplots. It also has extended ‘Describe contents’ for biplots and triplots as a help to their interpretation and many other small improvements.
Contribution biplots (Greenacre 2013a,b) are a solution to the issue that rare species are seemingly important as they typically appear at the margins of ordination diagrams, but are in fact not important at all. Contributions complement the fit statistics in Canoco.
Among the smaller improvements are the fit statistics in double constrained ordination and co-correspondence analysis and a handy overview of the most important changes since the Canoco 5 manual (see details).
Greenacre, M. 2013b. Contribution Biplots. Journal of Computational and Graphical Statistics 22: 107-122. http://dx.doi.org/10.1080/10618600.2012.702494
Canoco 5.1 implements new methods for trait-environment analysis and for micro-biome data analysis. Existing Canoco 5 customers can update to Canoco 5.1 with the usual update process from within Canoco 5.
The new method designed for micro-biome analysis is weighted log-ratio PCA (Greenacre and Lewi, 2009), and in similar vein, RDA. A fully worked example of the possibilities of Canoco for microbiome analysis is included in the new release.
Trait-environment analysis used to proceed via community weighted means (CWM) correlation or the fourth-corner correlation. These methods correlated a single trait to a single environmental variable. These analysis are now fully extended to the multi-trait multi-environmental variable case. The new method allows you to detect which traits show the highest correlation with the environment and, reversely, which environmental variables show the highest correlation with the traits. The method builds a regression model that allows you to quantify how much variation is explained by traits, by environment and by their combination.
The rationale for the new trait-environment methods has been summarized in a presentation and two published papers. The first paper links the fourth-corner GLM-based regression and gives the extension to the multi-trait multi-environmental variable case, which is simple double constrained correspondence analysis (dc-CA). The second paper gives a full description of dc-CA and the algorithm used by Canoco.
The linear-trait environment model of Cormont et al. (2011) has been extended similarly to double constrained principal component analysis (dc-PCA).
The new release has a reworked manual that comes with each new license. The free update comes with pdfs in the Canoco5/pdf folder containing the major changes in Canoco 5.1 (see details).
Copyright © 2021 上海卡贝信息技术有限公司 All rights reserved.