Effects of long-term drought on plant community structure in western Canadian peatlandsCourtney A. Miller, Brian W. Benscoter, and Merritt R. Turetsky

Department of Integrative Biology, University of Guelph, Canada

1. Introduction Warmer and drier climates may compromise the ability for peatlands to serve as a net carbon (C) sink by stimulating microbial activity and soil C losses. However, climate- and nutrient- mediated changes in plant community structure will influence net ecosystem productivity and will indirectly influence decomposition rates through organic matter quality inputs to soils (cf. Laiho et al. 2003).

Water table drawdown in peatlands can have either negative (Minkkinen and Laine 1998, Weltzin et al. 2000) or positive (Moore and Dalva 1993, Freeman et al. 1997) feedbacks to carbon emissions depending on site and vegetation characteristics. While lower water tables generally are expected to increase decomposition rates in the short-term, several longer-term studies have shown that drainage of Finnish peatlands ultimately increased soil C storage through the relatively slow response of increased inputs (cf. Minkinnen et al. 2002).

Western Canada is a region in which peatlands are expected to be most impacted by climate change, given the high rates of evapotranspiration relative to precipitation (Tarnocai et al. 2000). The goals of this research are to investigate the impacts of sustained drought in Alberta on peatland vegetation structure and productivity.

Figure 1. Images of the A) road-impacted bog (R55), B) experimentally-ditched McLennan fen, and C) road-impacted rich fen (R808), Site photos of the treatment plots (altered hydrology) are shown on top; aerial photographs of each site are shown at bottom.

Freeman C, Liska G, Ostle NJ, Lock MA, Hughes S, Reynolds B, Hudson J (1997). Enzymes and biogeochemical cycling in wetlands during a simulated drought . Biogeochemistry, 39, 177-187.Laiho R, Vasander H, Penttila T, Laine J (2003). Dynamics of plant-mediated organic matter and nutrient cycling following water-level drawdown in boreal peatlands. Global Biogeochemical Cycles, 17, 1053-1063.Minkkinen K, Laine J (1998). Long-term effect of forest drainage of the peat carbon stores of pine mires in Finland. Canadian Journal of Forest Research, 28, 78-186.Minkkinen K, Korhonen R, Savolainen I, Laine J (2002). Carbon balance and radiative forcing of Finnish peatlands 1900–2100 – the impact of forestry drainage. Global Change Biology, 8, 785-799.Moore T.R, Dalva M 1(997). Methane and carbon dioxide exchange potentials of peat soils in aerobic and anaerobic laboratory incubation. Soil Biology and Biochemistry, 29, 1157-1164.Tarnocai C, IM Kettles, Lacelle B (2000). Geological Survey of Canada, Open File 3834. Scale 1:6 500 000. Ottawa. Natural Resources Canada.Weltzin JF, Pastor J, Harth C, Bridgham SD, Updegraff K, Chapin CT (2000). Response of a bog and fen plant communities to warming and water-table manipulations. Ecology, 81, 3464-3478.

We examined the long-term effects of altered hydrology (via ditching and road-impacts) on plant community structure in several Alberta peatlands. Preliminary results from the McLennan experimentally-ditched fen were consistent with other studies that have showed increased tree productivity post-drainage (cf. Liaho et al. 2003; Minkkinen and Laine, 1998). However, results from the road-impacted sites were more complex. While the treatment plot at the R808 fen showed evidence of increased afforestation and higher total species richness relative to the control, there were few consistent differences in moisture content and vegetation structure between the treatment and control plots at the R55 bog. Our future work will investigate the effects of peatland type, peat structure, and nutrient loading on ecosystem responses to altered hydrology.

3. Results

Figure 4. Biodiversity index for the two treatments at each site. Data are means ± 1 SE.

Figure 2. Volumetric water content (mV) of surface peat at each site in mid-summer and early fall periods. Data are means ± 1 SE.

Our sites varied in dominant moss cover, with feather mosses dominating the McLennan fen, Sphagnum dominating the R55 site, and other true mosses (i.e., Polytrichum, Dicranumspp) dominating the R808 site (data not shown). Across sites, only Sphagnum cover varied between control and treatment plots, averaging 30% in control and 11% in treatment plots. Shrub cover varied by a site x plot interaction (Figure 3). Shrub cover did not differ among plots at the two road-impacted sites, but was higher in the control treatment at the McLennan site.

In each of the three sites, surface peat in the control plots was wetter than peat in the treatment plots (Figure 2). Densiometer data showed that the % of canopy cover was higher (more shading) in the treatment plot relative to the control at each site (data not shown).

5. References

Funding was provided by the Natural Sciences and Engineering Research Council of Canada. Many thanks to Mike Waddington, Mike Flannigan, Bill deGroot, and Eric Kasischke for comments on ths research, and to the Peatfire field crew, including S. A. Baisley, D. Thompson, and P. Van Hooren.

6. Acknowledgements

C.A. B.

4. Conclusions

We are investigating two proxies for long-term drought, including the i) the construction of logging roads through peatland complexes that led to the diversion of hydrologic flow, creating a drier downslope area within the peatland, and ii) a large-scale ditching experiment, initiated by the Canadian Forest Service and the Alberta Land and Forest Service in the 1980’s. This experiment involved the creation of a series of ditches in several peatlands located throughout north-central Alberta. Thus far, we have selected 3 study sites, including a road-impacted bog (R55 bog), a road-impacted rich fen (R808 fen), and an experimentally-ditched treed fen (McLennan fen).

In each site, we established a 100m2 treatment plot, as well as a 100m2 control plot located away from the hydrologically-altered area. Within each plot, three 0.25m2 quadrates were randomly located for measurements of species composition, canopy cover and available PAR (photosynthetically active radiation). Trees were harvested from a 100m2 plot in each plot (25m2 plot at the McLennan site). Additionally, we measured volumetric water content, soil temperature (in 5 cm increments from 0-30 cm), and water table depth several times during the growing season of 2009.

Percent cover data was analyzed using a 2-way ANOVA and Tukey post hoc comparison of means tests.

Site PlotTree Diameter

(mm)# of Trees/

m2

R55 Treatment 20.17 ± 3.6 0.33

Control 25.88 ± 3.7 0.57

McLennanTreatment 52.76 ± 9.7 4.4

Control 30.62 ± 6.3 1.32

R808Treatment 41.50 ± 7.8 0.59

Control 31.64 ± 7.6 0.57

Table 1. Diameter (dbh) and density of trees for each plot. Data are means ± 1 SE.

Figure 3. Percent shrub cover at each plot. Means with same letter superscripts are not different from one another.

CCC

B

AA

At the McLennan fen, the treatment plot had lower tree density but larger trees than the control plot (Table 1). Both the density and size of trees was higher in the treatment plot than the control at the road-impacted fen (R808). However, trees were smaller in the treatment plot at the road-impacted bog (R55) (Table 1).

Relative to the control plots, species richness was lower in the treatment plots at the two fen sites (McLennan and R808). This is due primarily to the loss of moss species in the treatment plots. However, at the R55 bog site, there was no difference in species richness between the control and treatment plots (Figure 4).

2. Methods