3.2 Heavy metal uptake and tolerance in S. cataractae: common garden experiment
Total and relative concentrations of Cd and Cu increased substantially in all treated plants compared to controls as shown in Figure 2 (χ²treatment = 4.57 for total Cd; χ²treatment = 143.07 for total Cu; χ²treatment = 40.88 for relative Cd in leaves; χ²treatment = 28.89 for relative Cd in stem; χ²treatment = 52.15 for relative Cu in leaves; χ²treatment = 102.36 for relative Cu in stem; p < 2.2e-16 in all cases). For Cu, all populations reached similar total concentrations as well as similar relative levels of Cu on the stems at the end of the experiment (Fig. 2A and C), except for treated Sc4 that accumulated significantly more Cu in the stems than the other populations. The magnitude of the treatment effect, however, differed among populations for the relative levels of Cu in leaves as indicated by the significant “Population by Treatment” interaction (χ²pop:treat = 8.61, p < 0.0001). In particular, populations Sc1 and Sc4, from the edges of the mine, showed higher relative levels of Cu in leaves after treatment than Sc2 and Sc3, from the center (z = -6.82 for Sc1 vs Sc2; z = -6.81 for Sc1 vs Sc3, 5.45 for Sc4 vs Sc2, 5.51 for Sc4 vs Sc3; p<0.0001 in all cases). We also found that the levels of Cu in the stems exceeded those in the leaves in Sc2, Sc3, and Sc4 (z = 5.23, p < 0.0001 for Sc2; z = 5.31, p < 0.0001 for Sc3; z = 2.50, p < 0.05 for Sc4) but not in Sc1 (z = -1.09, p = 0.405).
For Cd, Sc3 and Sc4 reached significantly higher total concentrations than Sc1 and Sc2 (Fig. 2D), and the relative levels in leaves showed a similar pattern, except that Sc3 did not differ significantly from Sc2 and Sc4 (Fig. 2E). All four populations reached similar relative levels of Cd in the stems. Finally, contrary to what we found for Cu, the levels of Cd in the leaves exceeded those in stems only for Sc4 (z = -4.384, p<0.001), and they did not differ among the other three populations.
Generalized linear models on growth revealed a significant “Population by Treatment” interaction in the effect of Cu in plant length, leaf length and leaf width (χ²pop:treat = 0.80, p<0.05 for plant length; χ²pop:treat = 0.57, p<0.01 for leaf length; and χ²pop:treat = 0.37, p<0.05 for leaf width). For Cd, this interaction was only significant for the length of the leaves (χ²pop:treat = 0.49, p<0.05). In general, gametophores from Sc2 were smaller than those from all other populations when grown in control conditions. Both metals seem to enhance Sc2 growth, as shown by the tendency towards higher means in all traits in treated plants compared to controls (means increased by 50, 14, 11% in Cu-treated plants for plant length, leaf length and leaf width respectively, and by 2, 14, and 11% in Cd-treated plants for plant length, leaf length and leaf width respectively), although the differences were statistically significant only for plant length between control and Cu-treated (z = 2.88, p<0.05). Populations Sc1 and Sc4 showed the opposite trend, with generally higher trait means in controls than in treated plants. Tests yielded significant differences for plant length between control and Cd-treated (z = -2.47, p<0.05; Fig. 3E), and leaf length and leaf width between control and Cu-treated in Sc1 (z = -3.36, p<0.01 and z = -2.94, p<0.05 respectively; Fig. 3B,C), and for leaf length between control and Cd- and Cu-treated plants in Sc4 (z = -3.01, p<0.05 and z = -2.79, p<0.05 respectively; Fig. 3B,F). Treatments did not significantly affect Sc3 and, overall, trait means of control and treated plants are pretty close.
Finally, Cu caused a significant and consistent increase in the concentrations of MDA in all populations (χ²treatment = 0.35, p < 2.2e-16; Fig. 3D) that averaged overall 14.2 µM in control cultures and 80.5 µM in Cu-enriched media. This effect was not significant for Cd-enriched cultures that exhibited mild and non-consistent changes in MDA concentrations (Fig. 3H).