# Canoco5 Support Site

Following errata apply to Canoco 5.1 manual printing that contains, on its second page, a copyright date range "1998-2018".

pg. 162: After Eq (5.1) and the first sentence ending with "covariate" add:

In double-constrained ordination (dc-CA and dc-PCA), the adjusted R2 is calculated by replacing the "(n-1)/(n-DF-1)" term in Equation (5.1) by "(n-1)(m-1)/((n-1)(m-1)-dfE*dfT)" where n is the number of cases, m the number of RV (typically, species), dfE the number of degrees of freedom of the predictors for cases, and dfT the number of degrees of freedom of the predictors for RV. Without multicollinearities and for numerical predictors, dfE is equal to the number of explanatory variables (environmental variables) and dfT is equal to the number of traits. The rationale for this definition is that (n-1)(m-1) is the full dimension of the interaction space between cases and RV which is modelled by dc-CA and dc-PCA by a space of dfE * dfT degrees of freedom. Note that the adjusted R2 so defined reduces to Equation (5.1) if dfT = m-1, i.e. if there are effectively no constraints on the RV-scores.

pg. 198: In Table 5.18, the reciprocal regression between the case and response variable (RV) scores hold true up to a constant of proportionality γ = √(n*m) or in the weighted case: √(TSS)/TAU  (defined in Eq (4.7) on pg. 131). Replace the first "=" by "= γ" in all lines with a case x* or RV score b on the left-hand side of the equation. The biplot approximations are correct.

pg. 201: In Table 5.20, similar changes as in Table 5.18 are needed. Replace the first "=" by "= γ" in all lines with an unconstrained case x* or RV score b on the left-hand side of the equation.

pg. 218: In Table 5.36, similar changes as in Table 5.20 are needed. Replace the first "=" by "= γ" in all lines with an unconstrained case x* or unconstrained RV score b* on the left-hand side of the equation.

pg. 224: In Eq. (5.11) and (5.12), replace the "=" by "= γ". See the erratum for pg. 198 above.

pg. 229: In Eq. (5.27) and (5.28), replace the "=" by "= γ". See the erratum for pg. 198 above.

pg. 301: In the paragraph before the last one, both occurrences of the word Moisture should be replaced by the word Management. In the last paragraph of the same page, the pair of words with partitioning on the second line should be replaced by without partitioning.

Following errata apply to Canoco 5 manual printing that contains, on its second page, a copyright date range "1998-2012".

pg. 23, 2nd line of 3rd paragraph: should now read "and check below the list that the chosen worksheets will define 2 project tables. After"...

pg. 24, 3rd paragraph: replace first sentence with the following three sentences: "Both data-tables are imported into a new Canoco 5 project and Canoco informs you that the Environment data table contains factors. This is not a problem report, rather additional information that you might even suppress for the next import event. After you close this box with OK, a dialog titled Introductory Analysis is shown."

pg. 25: figure at top has changed, so that the full page looks as follows.

pg. 32: figure in the middle part has changed, so that the full page looks as follows.

pg. 102: both figures have changed, so that the full page looks as follows.

pgs. 104-106: content of Tables 4.2 and 4.3 has changed, so the three pages look as follows.

pg. 115: figure has changed and the description below it was extended, so that the full page looks as follows.

pg. 116: first figure has changed and a new paragraph was inserted below the second figure, so that the full page looks as follows.

pg. 134: figure has changed and the full page looks as follows.

pg. 137: figure has changed and the full page looks as follows.

pg. 150: figure has changed as well as the text of next-to-last paragraph, so that the full page looks as follows.

pg. 163: second figure has changed, as well as the text of the paragraph that follows it, so that the full page looks as follows.

pg. 179: insert following paragraph before the one starting with "In linear methods":
"With more than 100 explanatory variables, the above correlation matrix is not inserted into the log view, unless this default behaviour is overridden (see section 7.3.3.1.8). "

pg. 288, 6th and 7th line of the last paragraph: change value 12 to 14 and value 43 to 22.

pg. 294: the figure has changed, so that the full page looks as follows.

pg. 297: the figure has changed, so that the full page looks as follows.

pg. 328: the figure has changed, so that the full page looks as follows.

pg. 349: the figure has changed, so that the full page looks as follows.

pg. 380: second figure has changed, so that the full page looks as follows.

pg. 382: after the 1st paragraph append following two:
"AIC with finite sample correction - when chosen, Canoco calculates (and uses during model selection) the AIC with finite sample correction, proposed by Hurvich & Tsai (1989).
Do not output correlation matrix - when checked, the matrix of correlations among explanatory or supplementary variables is not placed into log when there are >100 predictors."

pg. 394: replace second half of 2nd paragraph (i.e. the text starting with "Canoco 5 computes") with the following text:
"When the AIC with finite sample correction option is checked (see section 7.3.3.1.8), Canoco computes and uses AICc version of the AIC statistic, as proposed by Hurvich & Tsai (1989)."

pg. 413: the second figure and the first part of the paragraph below it has changed, so that the full page now looks as follows. Note that a part of the text now overflows into page 414, but there is no change in that text.

pg. 453: before the section 7.7 (Graph submenu) insert following text:
"7.6.11 Clone unlocked
This command is enabled only when the active analysis was created by the New Analysis Wizard, and creates a customizable copy that can be more freely modified."

pg. 460: Append following text at the end of first paragraph:
"Additionally, when you choose axis flipping in analysis options (see section 7.6.9), it is applied in XY(Z) diagram whenever an axis represents ordination scores."

pg. 487: after Hope, A.C.A. citation, insert following one:
"Hurvich, C.M. & Tsai, C.L. (1989) Regression and time series model selection in small samples. Biometrika, 76, 297-307."

pg. 493-496: several new entries were added to Index, so that it now looks as follows.