CAPdiscrim error: numeric 'envir' arg not of length one

Question

I'm trying to apply a CAP function to chemical data collected in different years.

I have a data archive:

head(ISPA_data)

  SrCa   MgCa  MnCa  RbCa  CuCa   ZnCa   BaCa  PbCa NaCa  LiCa  CoCa   NiCa
1 5178 25.101 9.334 0.166 4.869  8.379 34.846 0.194 5464 0.313 2.510 25.181
2 6017 22.922 7.185 0.166 4.685  8.720 24.659 0.154 4600 0.300 2.475 25.060
3 5628 26.232 6.248 0.179 4.628 10.157 23.942 0.166 5378 0.300 2.529 25.252
4 4769 35.598 7.683 0.131 4.370  8.735 50.068 0.180 5938 0.568 2.159 21.645
5 5330 28.284 6.828 0.130 5.370 12.742 34.257 0.220 5614 0.397 2.275 23.852
6 5786 24.603 4.797 0.156 5.317 13.331 66.896 0.117 5001 0.423 2.298 24.361

and a environmental dataset:

head(ISPA.env)

Year OM Code Location
  <dbl> <chr>   <chr>    <chr>   
1  1975 0.04349 CSP75_25 CSP     
2  1975 0.0433  CSP75_28 CSP     
3  1975 0.04553 CSP75_31 CSP     
4  1975 0.0439  CSP75_33 CSP     
5  1975 0.02998 CSP75_37 CSP     
6  1975 0.0246  CSP75_39 CSP 

When performing CAPdiscrim,

Ordination.model1 <- CAPdiscrim(ISPA_data~Year, 
                                ISPA.env, 
                                dist="euclidean", 
                                axes=4, 
                                m=0, 
                                add=FALSE, 
                                permutations=999)

this Error occurs:

Error in eval(predvars, data, env) :
numeric 'envir' arg not of length one

Besides: Warning message:

In cmdscale(distmatrix, k = nrow(x) - 1, eig = T, add = add) :
only 13 of the first 19 eigenvalues are > 0

All data has the same length.

Can anyone help me? Thanks!