Using High-Resolution Topographic Data to Document Large Wood Sources in a New England Watershed
Shortcut URL: https://serc.carleton.edu/59966
Location
Continent: North America
Country: United States
State/Province:Maine
City/Town:
UTM coordinates and datum: none
Setting
Climate Setting: Humid
Tectonic setting: Passive Margin
Type: Process
Show caption
Hide
Description
The presence of large woody debris (LWD) in streams has been associated with an array of ecological benefits. The rerouting of water around and over LWD (defined as pieces of trees > 10 cm diameter and 1 m long) often diversifies channels by creating pools, which fish use as in-channel habitat. Decomposing LWD can deliver valuable nutrients to the stream, enhancing habitat for the organisms that live in the channel. The zones of differing water velocities created by LWD causes sediment to be deposited in patches, creating distinct zones of substrate habitat for aquatic organisms.
Even though we have established that LWD serves an essential function in improving stream habitat, our understanding of how wood is delivered to channels remains limited. LWD can enter streams through many processes – for example, channel migration and bank failure, landsliding, simple tree mortality, and wind blow-down (Benda et al., 2003). While our knowledge of LWD delivery has increased with prior research, we know even less about how much LWD is available to streams. While counting riparian trees might provide an inventory of potential LWD that could get to channels, that method is labor-intensive and time-consuming, and can cover only a limited area.
An alternative that we have explored is to use high-resolution laser topographic data to offer a glimpse into the amount of LWD that is capable of reaching channels. Airborne light detection and ranging (or lidar for short) captures topographic data by recording the travel time of a laser pulse from an airplane to the ground. Because of the high frequency of laser pulses emitted, some strike the forest canopy, while others travel through the trees and strike the ground. As a result, two digital elevation models (or DEMs) can be produced: one of the tree-tops, and one of the bare earth below – and by differencing these two, a DEM of tree height throughout the lidar survey area can be produced (Figure 1).
Once the vegetation height DEM is produced, it can be restricted to those areas which might be able to contribute LWD to the channel – as trees that are located far away from the river will probably never contribute LWD. We can use the bare-earth DEM to delineate the valley width, which might deliver LWD to the channel through a number of mechanisms, including migration of the channel undercutting banks and downing trees and land sliding, for example. Using this valley-width-restricted vegetation height DEM, the amount of potential LWD along the course of an entire channel can be quantified (Figure 2). We used a simple cutoff to classify potential LWD: was the vegetation able to span the adjacent channel? Channel-spanning LWD has been shown to be especially important in promoting quality aquatic habitat – not only is it able to trap smaller pieces of wood, creating jams, but it also may be more stable in high flows, creating longer-lasting habitat improvement (Lienkaemper and Swanson, 1987).
We applied this method along the 76 km-long Narraguagus River in coastal Maine (USA), which has been intensively logged since the beginning of European colonization more than three centuries ago. The watershed is also one of eight Maine streams which host the last remaining wild runs of Atlantic Salmon (Salmo salar) in the United States (NRC 2004), and as a result, maintaining quality aquatic habitat and understanding what conditions in the channel might be like in the future are of importance to stream and watershed managers alike. Application of our lidar-derived LWD evaluation method reveals a few important trends in the distribution of LWD along the Narraguagus River. First, LWD is primarily concentrated in the upper parts of the stream – over 50% of all the channel-spanning LWD available to the stream is located along the uppermost three kilometers of river. More than 75% of all the available channel-spanning LWD is found in the upper 18 km of the river, with the remaining 51 km accounting for just 10% of the available channel-spanning LWD (Figure 3). This disproportionate LWD availability may be due to the wider channel in the downstream section which restricts the amount of LWD which is able to span the channel in these areas (Figure 4).
Our lidar-derived LWD evaluation method is advantageous to field inventories of potential LWD in that it is (A) more rapid and requires less labor, and (B) can cover spatially broad areas. Knowledge of where large stocks of potential LWD are located within watersheds will enable river managers and restoration practitioners to target areas for forest conservation and anthropogenic LWD additions, with the knowledge that their efforts may have a greater effect in these areas where natural processes can be complemented.
Even though we have established that LWD serves an essential function in improving stream habitat, our understanding of how wood is delivered to channels remains limited. LWD can enter streams through many processes – for example, channel migration and bank failure, landsliding, simple tree mortality, and wind blow-down (Benda et al., 2003). While our knowledge of LWD delivery has increased with prior research, we know even less about how much LWD is available to streams. While counting riparian trees might provide an inventory of potential LWD that could get to channels, that method is labor-intensive and time-consuming, and can cover only a limited area.
An alternative that we have explored is to use high-resolution laser topographic data to offer a glimpse into the amount of LWD that is capable of reaching channels. Airborne light detection and ranging (or lidar for short) captures topographic data by recording the travel time of a laser pulse from an airplane to the ground. Because of the high frequency of laser pulses emitted, some strike the forest canopy, while others travel through the trees and strike the ground. As a result, two digital elevation models (or DEMs) can be produced: one of the tree-tops, and one of the bare earth below – and by differencing these two, a DEM of tree height throughout the lidar survey area can be produced (Figure 1).
Once the vegetation height DEM is produced, it can be restricted to those areas which might be able to contribute LWD to the channel – as trees that are located far away from the river will probably never contribute LWD. We can use the bare-earth DEM to delineate the valley width, which might deliver LWD to the channel through a number of mechanisms, including migration of the channel undercutting banks and downing trees and land sliding, for example. Using this valley-width-restricted vegetation height DEM, the amount of potential LWD along the course of an entire channel can be quantified (Figure 2). We used a simple cutoff to classify potential LWD: was the vegetation able to span the adjacent channel? Channel-spanning LWD has been shown to be especially important in promoting quality aquatic habitat – not only is it able to trap smaller pieces of wood, creating jams, but it also may be more stable in high flows, creating longer-lasting habitat improvement (Lienkaemper and Swanson, 1987).
We applied this method along the 76 km-long Narraguagus River in coastal Maine (USA), which has been intensively logged since the beginning of European colonization more than three centuries ago. The watershed is also one of eight Maine streams which host the last remaining wild runs of Atlantic Salmon (Salmo salar) in the United States (NRC 2004), and as a result, maintaining quality aquatic habitat and understanding what conditions in the channel might be like in the future are of importance to stream and watershed managers alike. Application of our lidar-derived LWD evaluation method reveals a few important trends in the distribution of LWD along the Narraguagus River. First, LWD is primarily concentrated in the upper parts of the stream – over 50% of all the channel-spanning LWD available to the stream is located along the uppermost three kilometers of river. More than 75% of all the available channel-spanning LWD is found in the upper 18 km of the river, with the remaining 51 km accounting for just 10% of the available channel-spanning LWD (Figure 3). This disproportionate LWD availability may be due to the wider channel in the downstream section which restricts the amount of LWD which is able to span the channel in these areas (Figure 4).
Our lidar-derived LWD evaluation method is advantageous to field inventories of potential LWD in that it is (A) more rapid and requires less labor, and (B) can cover spatially broad areas. Knowledge of where large stocks of potential LWD are located within watersheds will enable river managers and restoration practitioners to target areas for forest conservation and anthropogenic LWD additions, with the knowledge that their efforts may have a greater effect in these areas where natural processes can be complemented.
Associated References
For further information:
- Kasprak A, Magilligan FJ, Nislow KH, Snyder NP. 2011. A lidar-derived evalutation of watershed-scale large woody debris sources and recruitment mechanisms: coastal Maine, USA. River Research and Applications. DOI: 10.1002/rra.1532
- Benda L, Miller D, Sias J, Martin D, Bilby R, Veldhuisen C, Dunne T. 2003. Wood recruitment processes and wood budgeting. In The Ecology and Management of Wood in World Rivers, Gregory S, Bowyer K, Gurnell A (eds). American Fisheries Society: Bethesda MD; 49-73.
- Lienkaemper GW, Swanson FJ. 1987. Dynamics of large woody debris in streams in old-growth Douglas fir forests. Canadian Journal of Forest Research 2: 150-156. DOI: 10.1139/x87-027.
- National Research Council. Committee on Atlantic Salmon in Maine. 2004. Atlantic Salmon in Main. The National Academies Press: Washington, DC.