adj] = 7, respectively), and neither did [TABULAR DATA FOR TABLE 3 OMITTED] they differ significantly between the VLEM plots on transect R and the HEM plots on transect P (t = 0.
Ordination of the 12 half-plots (and 24 quarter-plots) sampled on transect R, using all (n = 178) species, and either both basal area or density data revealed no separation LEM and VLEM plots on axes 1 and 2.
For transect R, there is little evidence for considering LEM and VLEM plots to be separate forest types on floristic grounds.
1 mm in the VLEM and LEM plots, respectively (t = 1.
The analogous design for transect R variables with six LEM and six VLEM plots sampled on two occasions (dates) led to 12 and 24 units for the two strata, respectively.
Therefore, soils from LEM and VLEM plots sampled in March 1990 on R were compared with soils sampled on LEM and HEM plots in March 1989 (designated P1) and in March 1990 (P2) on P.
Comparisons between concentrations in LEM and VLEM on transect R
In marked contrast to the transect P comparisons, those between VLEM and LEM plots on transect R showed very few significant differences (Table 14) and therefore only the means of the two forest types are shown.
Mean concentrations of (A) phosphorus and (B) nitrogen fractions ([[micro]gram]/g soil) in litter, root, and mineral layers of the 12 half-plots on transect R (six with very low abundances of ectomycorrhizal trees, VLEM; and six with low abundances, LEM) and the significance of the VLEM vs.
A posteriori comparisons among the six means were restricted to LEM on R with LEM on P1 and P2 (the cross-transect LEM comparison) and to VLEM on R with HEM on P1 and P2 (the extremes of forests sampled).
Comparing the VLEM plots on R with the HEM plots on P1 and P2 (Table 15) there was a consistent result across all layers, dates, and variables.