Results: A meta NMDS was used to help visualize the different disturbance treatments effect on the seedling performance data. The results of the seedling performance meta NMDS ordination (stress= 0.204) indicated a similar relationship to the one-way ANOVA tests in that seedling biomass, nitrogen and phosphorus appear as long vectors parallel to the ellipses that indicate different disturbance treatments on the ordination (Figure 6). This ordination also indicated that control sites had the highest biomass, nitrogen and phosphorus levels followed by Logging, fire, MPB and MPB plus logging disturbed sites (Figure 6).
Like the seedling performance ordination the fungal OTU ordination (stress= 0.204) indicated a similar relationship to the one-way ANOVA in that the length and direction of the vectors indicate a significant relationship between disturbance treatment and the fungal OTUs ECM1, ECM2, ECM3 and UNID1 (Figure 7). In addition to the meta NMDS, a heatmap was constructed using the fungal OTUs found to be significant following the one-way ANOVA test. Both the heat map and meta NMDS ordination indicated that ECM1 and ECM3 appear in higher concentrations in the control and logged sites followed by the fire sites and the MPB and MPB plus logging sites appear to have the lowest concentration of ECM1 and ECM3 (Figures 7 and 8). The heatmap and ordination indicated the opposite relationship for the fungal OTUs ECM2 and UNID1 in that these fungal groups were found in greatest concentration for the MPB and MPB plus logging treatments followed by the fire treatment and lastly the logging and control treatments with the lowest amount of ECM2 and UNID1 (Figures 7 and 8).
Like the seedling performance ordination the fungal OTU ordination (stress= 0.204) indicated a similar relationship to the one-way ANOVA in that the length and direction of the vectors indicate a significant relationship between disturbance treatment and the fungal OTUs ECM1, ECM2, ECM3 and UNID1 (Figure 7). In addition to the meta NMDS, a heatmap was constructed using the fungal OTUs found to be significant following the one-way ANOVA test. Both the heat map and meta NMDS ordination indicated that ECM1 and ECM3 appear in higher concentrations in the control and logged sites followed by the fire sites and the MPB and MPB plus logging sites appear to have the lowest concentration of ECM1 and ECM3 (Figures 7 and 8). The heatmap and ordination indicated the opposite relationship for the fungal OTUs ECM2 and UNID1 in that these fungal groups were found in greatest concentration for the MPB and MPB plus logging treatments followed by the fire treatment and lastly the logging and control treatments with the lowest amount of ECM2 and UNID1 (Figures 7 and 8).
Figure 6 (Below left): Meta NMDS ordination with seedling performance vectors plotted in dark blue and ellipses indicating different disturbance treatments. Figure 7 (Below Right): Meta NMDS ordination with fungal OTUs plotted as vectors in dark red and ellipses indicating different disturbance treatments.
Figure 8 (Below): Heatmap of fungal OTUs significant differences as a result of treatment.
Indirect gradient analysis was used to identify possible correlations between fugal OTUs and seedling height, biomass and nutrient composition. Upon running the NMDS ordination (stress= 0.196), two possible correlations were identified; one positive correlation where ECM1 and ECM3 may promote increased biomass and nutrient composition and a possible negative correlation in which ECM2 and UNID1 may decrease biomass and nutrient composition (Figure 9). In order to confirm that the correlations observed using the indirect gradient analysis are significant, a CANCOR constrained gradient analysis was run.
A CANCOR analysis was used to determine if any significant correlations existed between the allocation of carbon to fungal OTUs and seedling nutrient composition, height and biomass. Prior to running this analysis I checked my data and determined no unimodal relationships existed, therefore I was able to continue with the CANCOR analysis. The CANCOR analysis indicated four significant canonical functions Alpha level = 0.05 (Table 2). Although statistically significant, I will not focus on the last three canonical functions because these were correlations with fungal OTUs that were not significantly different as a result of inoculum treatment and as a result will not help to explain why the biomass, nitrogen and phosphorus content of seedlings differed between inoculum treatments. However, the results of the first canonical function indicated correlations with Fungal OTUs that were found to differ significantly with disturbance treatment. The first canonical function indicated a positive correlation between ECM1 and ECM2 with levels of seedling biomass, nitrogen and phosphorus content (Table 2). In addition, the first canonical function also indicated a negative correlation between ECM2 and seedling biomass, nitrogen and phosphorus (Table 2). These findings are similar to the correlations observed with indirect gradient analysis (Table 2, Figure 9).
A CANCOR analysis was used to determine if any significant correlations existed between the allocation of carbon to fungal OTUs and seedling nutrient composition, height and biomass. Prior to running this analysis I checked my data and determined no unimodal relationships existed, therefore I was able to continue with the CANCOR analysis. The CANCOR analysis indicated four significant canonical functions Alpha level = 0.05 (Table 2). Although statistically significant, I will not focus on the last three canonical functions because these were correlations with fungal OTUs that were not significantly different as a result of inoculum treatment and as a result will not help to explain why the biomass, nitrogen and phosphorus content of seedlings differed between inoculum treatments. However, the results of the first canonical function indicated correlations with Fungal OTUs that were found to differ significantly with disturbance treatment. The first canonical function indicated a positive correlation between ECM1 and ECM2 with levels of seedling biomass, nitrogen and phosphorus content (Table 2). In addition, the first canonical function also indicated a negative correlation between ECM2 and seedling biomass, nitrogen and phosphorus (Table 2). These findings are similar to the correlations observed with indirect gradient analysis (Table 2, Figure 9).
Figure 9 (Below Right): Indirect gradient analysis using meta NMDS ordination of biomass, nutrient composition and fungal OTUs. Biomass, nitrogen and phosphorus vectors are plotted in dark blue and fungal OTU vectors are plotted in dark red. Five disturbance treatments Control, Fire, Logging, MPB+Logging and MPB can be seen color coded on the graph.
Table 2 (Above Left): Table of the four significant canonical function as a result of the CANCOR constrained gradient analysis correlation values over 0.5 are in bold. P-values for each canonical function are located in the bottom row of the table.
Discussion: The results of the MANOVA, one-way ANOVA and Tukey adjusted pairwise comparisons indicate that disturbance altered fungal communities impact the nitrogen and phosphorus composition of seedlings, how seedlings accumulate biomass and how seedlings allocate carbon to the fungal OTUs ECM1, ECM2, ECM3 and UNID1 (Image 9). The differences as a result of disturbance treatment seen in the percent of carbon allocated to fungal OTUs is likely caused by the different fungal communities found at disturbed sites. Past work has demonstrated that changes in fungal community alter what fungal OTUs receive carbon from seedlings (Karst et al. 2015).
This work indicates fungal community composition alters seedling biomass, nitrogen and phosphorus levels and that seedling biomass, phosphorus and nitrogen are highest with the control treatment followed by logging, fire, MPB and MPB plus logging treatments. These results consistent with the expected results and are similar to observations made in the field where control sites had the highest amounts of regeneration followed by logging, fire, MPB and MPB plus logging. It is possible that the fungal communities found at control and logged sites are better suited to increase seedling biomass, nitrogen and phosphorus when compared to fire, MPB and MPB plus logged sites which due to the importance of increased nitrogen, phosphorus and biomass may result in greater regeneration success seen in the field.
This work indicates fungal community composition alters seedling biomass, nitrogen and phosphorus levels and that seedling biomass, phosphorus and nitrogen are highest with the control treatment followed by logging, fire, MPB and MPB plus logging treatments. These results consistent with the expected results and are similar to observations made in the field where control sites had the highest amounts of regeneration followed by logging, fire, MPB and MPB plus logging. It is possible that the fungal communities found at control and logged sites are better suited to increase seedling biomass, nitrogen and phosphorus when compared to fire, MPB and MPB plus logged sites which due to the importance of increased nitrogen, phosphorus and biomass may result in greater regeneration success seen in the field.
The results of both the indirect gradient analysis and CANCOR constrained gradient analysis indicate that seedlings may have increased biomass, nitrogen and phosphorus when grown on sites were fungal OTUs ECM1 and ECM3 are the dominant ECM fungal group present (Image 9). The results also indicate that growing seedlings on sites with relatively high concentrations ECM2 may result in decreased biomass, nitrogen and phosphorus (Image 9). Furthermore, these results stress the importance ECM fungal community composition as a factor in lodgepole pine seedling regeneration. These findings are supported by the literature which indicates that the composition of ECM can alter the ability of a host plan to acquire nutrients (Smith and Read 1997). Nitrogen and phosphorus are not only limiting nutrients for seedlings but the composition of ECM can also greatly impact the availability of these nutrients (Virginia et al. 1986, Smith and Read 1997). By improving the ability of the lodgepole pine seedlings to acquire growth limiting nutrients, fungal OTU groups ECM1 and ECM3 may enable seedlings to attain increased biomass.
This work emphasizes that land managers need to consider soil fungal community composition when attempting to regenerate lodgepole pine on disturbed sites, particularly the ECM soil fungal community. Additionally, this work may help inform the amendment of soils particularly for sites impacted by MPB and MPB plus logging to improve the biomass, nitrogen and phosphorus content of seedlings. Future work should investigate the viability of amending soils to encourage ECM1 and ECM3 and discourage ECM2. If viable, soil amendments could prove to be an important tool in improving the regeneration success of lodgepole pine following disturbance.
This work emphasizes that land managers need to consider soil fungal community composition when attempting to regenerate lodgepole pine on disturbed sites, particularly the ECM soil fungal community. Additionally, this work may help inform the amendment of soils particularly for sites impacted by MPB and MPB plus logging to improve the biomass, nitrogen and phosphorus content of seedlings. Future work should investigate the viability of amending soils to encourage ECM1 and ECM3 and discourage ECM2. If viable, soil amendments could prove to be an important tool in improving the regeneration success of lodgepole pine following disturbance.
Image 9 (Below): The results of this study suggest that fungal community composition is an important factor in the regeneration of lodgepole pine on disturbed sites.