Results & Discussion

Residual glyphosate declined drastically from year one to year two after treatment (Figure 11). Average glyphosate levels were compared to 0.1 ppm because this is the maximum residue limit set by the Government of Canada for human food consumption (Health Canada, 2012; Botten et al, 2021; Kolakowski et al, 2020). Throughout each of the subregions, the average glyphosate levels remained below 0.1 ppm (Figure 11, Table 6). The lower foothills, upper foothills, and lower boreal highlands regions had the highest residual concentrations during the second year after treatment relative to the other three subregions (Figure 11). The lower and upper foothills subregions are characterized by short summers and long cold winters. This could play a role in the slower degradation of glyphosate, as microbial activity is slowed and plants remain dormant (Botten et al, 2021; Helander et al, 2012). However, by the second year after initial treatment, all subregions tested had minimal residual glyphosate, with less than a 1% chance of glyphosate being over 0.1 ppm in the second year (Table 6). 

Average glyphosate within individual species followed a similar pattern of decline over the two years after treatment (Figure 12). Alder species (Alnus spp.) contained the highest glyphosate residuals initially but declined to levels relatively similar to the other species at year two. Residual glyphosate concentrations for species were also compared to the 0.1 ppm reference level (Health Canada 2012) (Table 7). The majority of the species tested had very minimal residual glyphosate after the second year, with less than a 1% chance that glyphosate would be above 0.1 ppm (Table 7). There are two exceptions to this finding, dwarf blueberry (Vaccinium caespitosum) and Canadian gooseberry (Ribes oxyacanthoides) both contained relatively high amounts of glyphosate at the two year point after treatment (Table 7). Dwarf blueberry has around a 70% chance of containing greater than 0.1 ppm of glyphosate after two years, and Canadian gooseberry has a 27% chance (Table 7). These results are similar to the results reported by Botten et al. (2021) and Wood (2019), which both observed low levels of residual glyphosate in shoot tissues one year after application.  

The overall plant community composition did not change too drastically over the two years after herbicide application (Figure 13). The ellipses of the PCoA have a fair bit of overlap, and the year 2 community composition appears to be shifting back towards a similar composition to that of the untreated reference plots (Figure 13). 

Some notable changes between the plant communities include the initial decrease of rose (Rosa acicularis), raspberry (Rubus idaeus), fireweed (Chamerion angustifolium), aspen (Populus tremuloides), highbush cranberry (Viburnum edule), alder (Alnus spp.), Canadian gooseberry (Ribes oxyacanthoides), and bracted honeysuckle (Lonicera involucrata). These species are all generally larger in form and have greater leaf area than low-lying and trailing plants. With the addition of GBH in year one, many of these species will have experienced a reduction in growth, reducing competition and allowing the understory species such as dwarf blueberry (Vaccinium caespitosum), velvet blueberry (Vaccinium myrtilloides), lingonberry (Vaccinium vitis-idaea), and kinnikinnick (Arctostaphylos uva-ursi) to cover a greater area; this can be seen in the shifts of the plant community down and to the left of the untreated reference plots (Figure 13).

Figure 13. PCoA of plant communities in three subregions of Alberta. Colored ellipses represent the overall shift in the plant community, comparing untreated plots (green), 1 year after treatment (yellow), and 2 years after treatment (red dashed). Vectors overlaid indicate species of interest (black), as well as environmental factors (grey) and Shannon's and Simpson's indices (orange). "CM" = central mixedwood. "LF" = lower foothills. "UF" = upper foothills. Only subregions with corresponding untreated reference plots are included in the ordination. 

As stated above, many of the plots are dominated by a few species. Relative frequency of fireweed, rose, raspberry, willow (Salix spp.), and blueberry species is much higher than that of the other species of interest (Figure 14). Frequency is measured here by the presence or absense of species throughout each of the natural subregions. Many species were only recorded a few times throughout all of the cut blocks, including alder, kinnikinnick, paper birch (Betula papyrifera), bracted honeysuckle, skunk currant (Ribes glandulosum), and mountain ash (Sorbus scopulina).  

Compared to the untreated cut blocks, several species of interest appear more frequently in the years following glyphosate application. In the central mixedwoods region, kinnikinnick, fireweed, rose, and lingonberry appear more frequently (Figure 14). The lower foothills also had more kinnikinnick, as well as increases in frequency of marsh labrador tea (Ledum palustre), velvet blueberry, and lingonberry. The upper foothills saw increases in kinnikinnick, marsh labrador tea, velvet blueberry, lingonberry, and highbush cranberry. As previously mentioned, species such as kinnikinick, lingonberry, and blueberries will appear more frequently as competition from overstory species is reduced. 

After initial application of GBH, relative abundance decreased for some species (Figure 15). Relative abundance was calculated using the % cover data, indicating how much area each plant takes up in each subregion. For species such as aspen, and raspberry, the abundance decreased following herbicide treatment. Raspberry and rose often occur together in similar ecosite types, and where raspberry is reduced, abundance of rose can be seen to increase (Figure 15). 

Again, abundance of kinnikinnick, lingonberry, dwarf blueberry, and velvet blueberry can be seen increasing as competition from other species is reduced.