AGU Fall 2011 Poster presentation

C.A. Miller1, M.R. Turetsky1, B.W. Benscoter2 1 Dept. Of Integrative Biology, University of Guelph, Guelph, Ontario, Canada 2 Dept. of Biological Science, Florida Atlantic University, Davie, Florida, U.S.A.

([email protected], [email protected], [email protected])

The effect of long-term drainage on vegetation structure and productivity in boreal peatlands

5.  Conclusions

7. Acknowledgements

6.  References

We measured aboveground biomass (vascular and moss) at the reference and drained plots in each site (excluding RPF2). Our results showed no significant differences in understory vascular biomass between treatment and control plots at any of the treed sites. However, shrub biomass increased with drainage at the open poor fen (ROF). Tree biomass increased

We measured aboveground productivity (NPP) at the McLennan and RMF sites only. At the McLennan fen, drainage led to a large increase in tree NPP and a small reduction in moss NPP (Fig 6,7). At the RMF site, drainage increased moss NPP and resulted in no change in tree NPP (Fig 7). Together, these results may indicate a negative relationship between moss and tree NPP.

4. How does long-term drainage influence plant biomass and NPP?

Poster Number: B23C-0439

ET ET

Fig 8._Schematic of potential positive feedback between tree biomass, evapotranspiration, and water table (WT) drawdown in peatlands.

3. How does long-term drainage influence plant species composition?

structure in the treed poor fens (McLennan, RPF1, RPF2). In these sites, drainage caused increases in feather moss and decreases in Sphagnum abundance (Fig 3). The RMF, RPF2, and RB2 sites showed a drainage x microtopography interaction, where hollows showed more changes in species composition than hummocks (Fig 4).

Fig 9. The 2011 wildfires in Alberta burned many drained peatlands (credit: Alberta Wildfire Info.)

2.  Study Sites

Site Description Rod-impacted Moderate Fen (RMF) Forested moderate fen Road-impacted Open Fen (ROF) Unforested poor fen Road-impacted Poor Fen 1 (RPF1) Forested poor fen Road-impacted Poor Fen 2 (RPF2) Forested poor fen Road-impacted Bog 1 (RB1) Forested bog Road-impacted Bog 2 (RB2) Forested bog McLennan ditched fen (McL) Forested poor fen

Table 1. List of drained peatland sites in north central Alberta, Canada.

Fig 1. Air photos of a A) road-impacted bog and B) experimentally ditched fen in Alberta. Arrows indicate location of treatment area. Photos on top are pre-drainage; photos on bottom are post-drainage.

A 1965

2001

B 1983

1999 Fig 7. Moss NPP and tree ANPP at the McLennan (triangle) and RMF (circle) sites Error bars are ±1 S.E. of the mean.

influence C storage by increasing woody debris inputs to soils. However, increases in woody biomass and/or decreases in Sphagnum also will increase fire risk in peatlands (Fig 8). Turetsky et al. (2011) found that drainage of a treed fen increased burn severity 9-fold during wildfire despite increases in peat accumulation following drainage. Thus, relationships between drying, succession and periodic disturbances (fire, insect outbreaks) are important for predicting the fate of peatland C pools.

1.  Introduction

We determined understory species composition at randomly located 0.25m2 quadrates in the treatment and control plots at each site (n=12-20/site). Data were analyzed using non-metric multidimensional scaling (NMDS) and multi-response permutation procedures (MRPP) with an α = 0.05. These analyses showed a significant effect of drainage on community

We selected six road-impacted sites and one experimentally ditched site in north central Alberta, Canada (Fig 1). At each site, we established a treatment (hydrologically altered) and control (pristine) plot (Fig 2; Table 1).

Fig 2. Images of the A) McLennan, B) RMF and C) RPF1. Site photos of the control plot are shown on the top; photos of the treatment plot are shown at the bottom.

A

B

A B C

Our results suggest that changes in vegetation with long-term drainage could .

Generally, drainage increased woody biomass, and decreased Sphagnum cover at most of the treed fen sites. Increased tree density and/or productivity is likely to stimulate water loss through evapotranspiration and interception (Sarkkola et al., 2010), which could facilitate further drying (Fig 7).

Minkkinen et al. (2002). Global Change Biology, 8, 785-799. Sarkkola et al. (2010). Canadian Journal of Forest Research, 40, 1485-1496. Turetsky et al. (2011). Nature Communications, 2, 514.

Funding was provided by the Natural Sciences and Engineering Research Council of Canada. Many thanks to Mike Waddington, Mike Flannigan, Mike Wotton, Bill deGroot, and Eric Kasischke for comments on this research, and to the Peat Fire field crew, including S. Andrew Baisley, Dan Greenacre, Tom Schiks, Abra Martin, Katarina Neufeld & James Sherwood.

Water table drawdown in peatlands can have either negative or positive effects on radiative forcing. While drier conditions are likely to stimulate decomposition of peat, long-term drainage of Finnish peatlands increased aboveground biomass and soil carbon pools through increased tree production and inputs to soils (Minkkinen et al., 2002). Understanding the influence of drying on peatland succession is necessary for understanding the effects of climate change and land-use on peat accumulation and potential C release. Here, we quantified changes in vegetation in peatlands impacted by several decades of drainage resulting from road construction. We also investigated vegetation changes that occurred with experimental ditching initiated in the mid 1980’s.

Drainage

Fig 6. At the McLennan site, tree ring width averaged 0.4 ± 0.02mm for the 10 years prior to drainage and 1.6 ± 0.10mm for the most recent 10 years.

(1986)

Fig 4. Results from the RMF site NMDS depicting differences in understory species composition with drainage (final stress = 5.66838, n = 12 with 23 taxa). While hummock vegetation was similar across the control and treatment plots, hollow species composition diverged between plots.

Axi

s 2

Axis 1

Control Hummock

Control Hollow

Treatment Hummock

Treatment Hollow

with drainage at all three treed poor fens (RPF1, RPF2, and McL) and one bog (RB2; Fig 5).

Fig 5. Change in aboveground tree biomass with drainage at each site * denotes significant differences between treatment and control areas within a site.

0

200

400

600

800

0 250 500

Tree

NPP

(g/m

2 /yr)

Moss NPP (g/m2/yr)

Treatment

Control Control Treatment

Fig 3. Change in moss abundance between treatment and control plots at each site. Data are means ± 1 S.E. Note – the RMF was dominated by true moss (Tomenthypnum and Drepanocladus spp) and showed no change in abundance.

WT WT

-100 0 100 Change in % cover of moss

Sphagnum

Feather Moss

RB2

RB1

RPF2

RPF1

McL

ROF

RMF

-400 -200 0 200 400 Change in aboveground tree biomass (kg/100m2)

RB2*

RB1

RPF2*

RPF1*

McL*

RMF