Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Beaver-generated disturbance extends beyond active dam sites to enhance stream morphodynamics and riparian plant recruitment


Given the direct effects of their dams on hydrology, sediment storage, and vegetation, beaver are widely acknowledged as ecosystem engineers. Here we explore the effects of beaver activity on channel processes and riparian plant recruitment beyond those dams and after dam abandonment in southwestern Montana, USA. Willow cuttings from beaver herbivory are commonly deposited along point bars, adding roughness and promoting sediment accumulation. Most cuttings are found <1 km downstream of an active dam. These cuttings often sprout, aiding in willow colonization and bar stabilization. Thirty-four radiocarbon ages show that beaver cuttings have accumulated by similar processes over thousands of years, adding to floodplain carbon storage. Breached dams can initiate meanders, increasing channel and riparian habitat diversity. Beaver activity thus generates a cycle of frequent disturbance, from dam building and riparian plant browsing through dam failure and abandonment, with each phase influencing channel and floodplain evolution and riparian plant recruitment.


Riparian ecosystems are coming under increasing stress from changing climate, increasing demands on water, and loss of flood-induced disturbance1. In addition, given the strong influence of beaver on fluvial and riparian function, and their decline across much of their native Northern Hemisphere habitat2, these ecosystem engineers have become an important area of focus in riparian science. Attention has centered on the effects of intact dams3,4 and direct impacts of herbivory5,6. Nonetheless, there has been limited investigation of the effects of beaver activity on stream reaches and riparian vegetation beyond their iconic dams and ponds, or on the continuing impacts of breached and abandoned dams. For example, beaver are notorious herbivores, generating a steady supply of beaver-chewed wood7 that is often transported downstream. Although seed dispersal may be the dominant and critical form of reproduction for riparian plants, sprouting from stem fragments can also play an important role in riparian plant reproduction, particularly for species that are adapted to floodplain disturbance8. In prior research along beaver-occupied streams in southwestern Montana and the Rocky Mountain region, we observed that willow cuttings produced by beaver herbivory and dam building are commonly transported downstream and accumulate along channel margins. Thus, we asked whether beaver-generated willow cuttings are deposited and sprout in sufficient quantities to facilitate riparian colonization in reaches below active beaver dams; whether cuttings and sprouts promote further sediment deposition; and if cuttings deposition has been a widespread and long-continued process, which can be addressed through identification and radiocarbon dating of subfossil willow cutting accumulations. In addition, we saw that beaver dam remnants often persisted for some years after breaching and abandonment, with the potential to influence channel morphology and evolution. Here, we document the continuing effects of remnant dams to consider the impacts of beaver beyond the life of active dams and ponds.

The entire network of a beaver occupied stream appears to be influenced by their presence, as dam sites cycle through being maintained, abandoned and ultimately failing. We have focused our work on three streams in the upper headwaters of the Missouri River System in southwestern Montana (Table 1, Fig. 1) that all feature unconfined meandering alluvial channels9, where active beaver colonies exist within protected landscapes with limited human impacts. On the three study streams, we document and quantify beaver cutting dispersal and additional less-direct effects of beaver activity within the dam-building and abandonment cycle, including interactions with channel processes.

Table 1 Characteristics of study streams.
Figure 1

Inset map (ESRI ArcGIS Online, © 2007 National Geographic Society, i-cubed) shows the study streams in blue with bold box showing location of main map. Main lidar shaded relief map ( shows sites with beaver cutting deposits along Odell Creek. Radiocarbon ages for the oldest beaver cuttings sampled at each site are shown as the median of the calibrated age probability distribution57, rounded to the nearest 10 years; uncertainties and measured radiocarbon ages can be found at:


Accumulations of willow cuttings from beaver herbivory

The feeding activities of beaver and the construction, maintenance, and breaching of beaver dams provide a steady supply of fine woody debris to the channel. Beaver-cut willow stems (hereafter termed cuttings) are generally small in diameter (1–3 cm), with a mean length of 10.4 cm (SE ± 0.22). The cuttings are deposited in clusters along the margins of the study streams and cover large areas of the point bars (Figs 2, 3). At 90 sites along Odell Creek, we recorded a summed total length of 228 m of beaver cuttings within 270 0.5 × 0.5 m quadrats. Beaver cuttings accumulated on 81% of point bar sites and 51% of all surveyed sites along Odell Creek, demonstrating the ubiquitous character of this process. Red Rock Creek and East Fork of Blacktail Deer Creek (EFBD Creek) showed similar accumulations of cuttings on point bars in meandering reaches (Fig. 3).

Figure 2

Beaver generation of willow cuttings (plant propagules) and distribution. 1. Willow food cache on East Fork Blacktail Deer Creek (1a); beaver cut willow for dams on Odell Creek (1b) 2. Some cuttings from loss or dam breaching are transported downstream. The dashed line shows the thalweg, i.e. the deepest part of the channel. The cuttings are eventually deposited in areas of low velocity and shear stress, often on the downstream margin of point bars, along with fine sediment and willow seeds. 3. The cuttings add roughness contributing to further fine sediment accumulation. 4. Some of the cuttings develop adventitious roots and sprout on the wet substrate of bars. 5. Sprouted willows grow into mature plants, increasing roughness and promoting point-bar growth and meander development. All willows in this Odell Creek photograph are growing on point bars, whereas the cutbank in the background is formed in a higher fluvial terrace with grassy vegetation. The youngest willows lie near the toe of the bar, coincident with the youngest sediment (see also (2) in this figure).

Figure 3

Box plot showing percent of 0.5 × 0.5 m quadrats within each reach (45 quadrats per reach), that contain willow cuttings produced by beaver herbivory along Odell Creek (OC, mean gradient = 0.004 m/m, mean sinuosity = 2.5, 3 reaches), Red Rock Creek (RRC, mean gradient = 0.003 m/m, mean sinuosity = 2.2, 3 reaches) and East Fork of Blacktail Deer Creek (EFBDC, mean gradient = 0.008 m/m, mean sinuosity = 1.8, 5 reaches), Beaverhead County, Montana. Thick lines show medians; x symbols indicate means; box limits denote 25th and 75th percentiles; whiskers extend 1.5 times the interquartile ranges; there are no outliers; medians are included in calculations of quartiles.

Controls on beaver cutting accumulations

To understand controls on beaver-cutting deposition, we investigated variability in accumulation along the margins of Odell Creek, Red Rock Creek and EFBD Creek. Below, we report the results of linear mixed effects analyses10, where Chi-squared likelihood ratio tests were used to assess how each parameter affected mean willow cutting length at sites as: χ2 (degrees of freedom) = test statistic, p = p-value. Where appropriate, we also include the mean and standard error of the mean (SE).

As beaver harvest willow for dam construction, food caches, and immediate consumption near active dam sites, these sites were hypothesized to be the most important parameter influencing total willow cutting length. Model results show that the distance downstream from an active dam significantly explains variability in total cutting length (χ2 (1) = 4.49, p = 0.03) (Fig. 4), and that total cutting length decreases by 0.06 cm/m (SE ± 0.03 cm/m) downstream of a dam site. We also hypothesized that the total number of dams upstream of a site would be important in the total cutting length, but this was not the case, as the farthest downstream site, with 7 dams on the stream above it, had the lowest total cutting length. Even the most upstream reach, with only 2 dams above it, had a greater total cutting length. Most cuttings are found less than ~1 km downstream of an active dam. These findings suggest that the mean travel distance of cuttings between sources and depositional sites is relatively low, and that dams efficiently trap cuttings from upstream.

Figure 4

Summed length of willow cuttings produced by beaver herbivory relative to potential controlling factors along Odell Creek, Montana. (a) Summed cutting lengths by site plotted versus site distance to nearest upstream dam. All sites are shown. (b) Summed cutting lengths for each channel morphological class across all sites. (c) Box plots showing summed cutting length across 0.5 × 0.5 m quadrats in each sediment class; center lines show medians; x symbols indicate means; box limits denote 25th and 75th percentiles; whiskers extend 1.5 times the interquartile ranges; outliers are represented by points; medians are included in calculations of quartiles. (d) Box plots showing summed cutting length across quadrats in each distance class (0, 1 and 3 m), where distances are from low-flow water edge. Box symbols and calculations are the same as in c.

Channel morphology strongly controls deposition of willow cuttings, creating distinct populations of cutting accumulations among point bars, straight reaches, and cutbanks (χ2 (2) = 6.41, p = 0.04), where point bars accumulate the greatest cutting length (mean = 202.8 cm (SE ± 33.2)), and cutbanks the least (Fig. 4). Because point bars were dominant locations of accumulation, we subdivided point bar sampling sites by upstream, midpoint and downstream locations. Point bar location was important in explaining much of the variability in total cutting length between sites (χ2 (2) = 770.13, p < 0.001), with downstream bar sites having the greatest cutting length (mean = 278.9 cm (SE ± 59.1 cm)), followed by upstream bar (mean = 235.4 cm (SE ± 65.1 cm)) and midpoint bar sites (mean = 126.7 cm (SE ± 54.9 cm)). Throughout a bend, the point of maximum shear stress moves from inside to outside of the bend. Greater cutting accumulation on the downstream edges of bars is consistent with low boundary shear stress in meander-bend flow11.

Without directly measuring shear stress, sediment size provides a way to understand the relative shear stress of cutting accumulation sites. The expected variations in shear stress around meander bends is reflected in grain size distributions, where the coarsest sediments are found in pools near the outside bank, just downstream of the point bar, while fine sediments are transported toward the inside and downstream part of the bend, where shear stresses are lower11. Sites with lower shear stress are sites that may protect seedlings from scouring flows. Grain size distributions are not only important for understanding shear stress, but also for determining whether cuttings are being deposited in sandy sediment that retains moisture needed for willow establishment12. Fine and medium sand were most common at downstream point bar quadrats at 1 m from the low-flow channel (n = 10 and 11 respectively) and were associated with large cutting accumulations (Fig. 4), although medium sand quadrats contained the greatest cutting accumulations across all distances from the channel (Fig. 4).

Location relative to the active channel explains a significant amount of the variability in total length of cuttings between quadrats (χ2 (2) = 618.6, p < 0.001) (Fig. 4). Within each site, quadrats were placed at the low-flow stream edge and at 1 m and 3 m from the stream edge. The 1 m quadrats accumulated the greatest total length (mean = 323.8 cm (SE ± 55.62 cm)), followed by the quadrats adjacent to the low flow channel (0 m quadrats, mean = 247.3 cm (SE ± 51.1 cm)). The 3 m quadrats were the least likely to accumulate cuttings (mean = 164.2 cm (SE ± 70.9 cm)). In most cases, sites at the 3 m distance were on average 70 cm above the low-flow channel elevation and 40 cm above the bankfull channel elevation, and only larger floods are likely to push cuttings this far onto the floodplain.

Channel sinuosity, morphology and hydraulics appears to control cutting accumulation as well. Along Odell Creek, the upper reaches with steeper gradient and coarse gravelly bed sediment have no major accumulations of beaver cuttings. Field observations also indicate that the lower gradient, meandering reaches of Odell Creek and Red Rock Creek accumulate more willow cuttings than the higher gradient, less sinuous reaches of EFBD Creek (Table 1, Fig. 3). On small headwater streams such as Alkali Creek (Table 1), the lack of well-developed point bars appears to inhibit beaver cutting accumulations, even though consecutive beaver occupancy surveys indicate that dams may persist there for at least a decade. On larger regional rivers such as the Madison and Clark Fork Rivers in Montana13, beaver have currently and historically occupied smaller side channels; these exhibit some beaver cutting accumulations, but the cuttings appear to have limited influence on river processes in these high-energy systems.

Beaver cuttings and willow recruitment

We counted sprouts from cuttings in each quadrat, which yielded a mean number of sprouts per quadrat of 0.5 (SE ± 0.1), with a total of 72 sprouts across all quadrats in the study area. Although this number is relatively low compared to the total length of cuttings, sprouts were present at 25% of all sites and appeared to be more numerous with a greater total cutting length (Fig. 4). Sprouts were most commonly associated with sites with sandy deposits, especially medium sand. Forty-five percent of medium sand quadrats had sprouts, with a mean of 1.2 sprouts per quadrat (SE ± 0.4), possibly because of significantly more cutting length available for sprouting (7620 cm). Coarse sand quadrats had the highest mean sprout content (1.6 sprouts/quadrat (SE ± 1.4) and showed the greatest sprouting success per available cutting length (1020 cm). Overall, these data indicate that beaver herbivory results in deposition of abundant cuttings on suitable substrates, making vegetative propagation a viable mechanism for recruitment. We tested whether total cutting length at a site had a strong control on the number of sprouts produced and found that the relationship was not significant.

The effect of beaver cuttings on channel processes, floodplain evolution and carbon burial

Cuttings added to stream channels by beaver activity are also likely to alter sedimentation processes at low-energy depositional sites. Although the cuttings are small in diameter, they are deposited in clusters (Fig. 2) with the initial deposition of fine woody debris making it more probable that additional cuttings will be trapped at the site14.

The abundance of beaver cuttings in older Holocene-age deposits along Odell Creek and Red Rock Creek demonstrates that beaver cuttings accumulating on point bars is ubiquitous and has been throughout the Holocene. Beaver cuttings, preserved for millennia, are evidence of relatively rapid burial and demonstrate that beaver-inhabited headwater alluvial streams are sites for long-term storage of detrital organic matter (Figs 1, 2, 5)9. Holocene terraces from 1.2–3 m above the low-flow channel represent former floodplains of these streams, and their deposits are commonly well-exposed in cutbanks and well-distributed throughout the meandering middle reaches of the creeks. Thirty-four radiocarbon (14C) ages show that beaver cuttings range in age from ~6030–380 cal yr BP (Fig. 1) demonstrating that deposition of beaver cuttings has been a common process over millennia. The subfossil beaver cuttings appear in discrete, laterally extensive layers with high concentrations of willow stems, cleanly cut at an angle typical of beaver-chewed wood (Fig. 5). The cuttings layers have a narrow range in radiocarbon age and are mostly contained within sandy sediment overlying coarser sand, pebbles, or gravel; on top of the cuttings are fine sands grading to thick deposits of sandy silt. This upward-fining trend is typical of point-bar deposits and overlying overbank sediments15. Sedimentary sequences associated with abandoned beaver dams in the study region feature localized woody debris accumulations, berm-like surface forms, and thick, muddy, organic-rich pond sediments16. In contrast, we find that Holocene sediments containing beaver cuttings along the meandering streams of this study most closely resemble point-bar deposits. The abundance of beaver cuttings on modern point bars (Fig. 4(b)), including those being buried by sediment, provides a strong modern analog in support of this interpretation.

Figure 5

(a) A composite stratigraphic column showing the typical context of beaver willow cuttings buried in Holocene sediments along Odell Creek. Units observed at most sites and their mean thickness and grain size are shown. Cartoon sticks in the second unit from the bottom show the typical position of cutting deposits. (b) Photographs show examples of in situ cuttings at 2 sites, where the top photo shows cuttings exposed below the water surface; median calibrated radiocarbon ages57 for collected samples, rounded to the nearest 10 years, are noted. White ellipses highlight the cuttings in each image. (c) Plot shows median calibrated radiocarbon ages and depths of collection for site 12R35 – also shown in lower photograph in (b); “c” represents charcoal and “b” represents beaver cutting samples. The circle highlights the beaver cutting ages and depths.

Active versus abandoned dams

Beaver dams in southwest Montana streams have variable longevity as intact structures, but the breached remnants of those dams persist much longer and make up the majority of instream beaver structures. In the field, abandoned dams were identified where some remnant of the dam lay within the main channel. In summer 2009, 10 km of Odell Creek contained 9 active dams and 21 abandoned dams. In surveys of 10 km of EFBD Creek in 2015, 24 active dams and 33 abandoned dams were identified. On both Odell Creek and EFBD Creek, long-term monitoring shows that beaver reoccupy sites at >5 year intervals, with locations of beaver occupancy shifting throughout the available habitat along the streams. Dam sites are active over an average of three years on these two midsize streams. We measured the upstream influence on water surface slope on Odell Creek and EFBD Creek and found that active dams are affecting slope over 3% and 4%, respectively, of the total surveyed channel length. On smaller, adjacent Alkali Creek, a 2016 survey identified 24 active dams and 24 abandoned dam sites in 10 km, and comparison with a 2009 survey and USDA NAIP imagery from 1995–2014 show that dam positions for 6 of the largest dams have remained consistent for 7 years. One of the large beaver wetlands on Alkali Creek has remained intact since 2005, with others persisting for >5 years. On Alkali Creek, abandoned dam sites maintained an average of 70% of cross-channel extent. On Odell Creek and EFBD Creek, some dam remnants can persist within channels and influence flow patterns and sedimentation for at least a decade.

Along the study steams, we observed that meanders had been initiated by beaver dam remnants that deflected flow toward the opposite bank17. Dam remnants that were stabilized by live plant stems sprouted from cut willows used in dam construction were particularly effective in directing flow. In typical meander bends, shear stress is concentrated at the outer bank, promoting cutbank erosion and the transport of coarse bed load, whereas fine sediment is deposited along the inner bank11. In the case of a breached dam, a remnant can act as the inner bank, with deposition further enhanced by sprouting willow. Thus, the beaver dam remnant forces the channel to shift laterally over time.

As beaver occupy and abandon sites, dam sites shift from intact to breached, to abandoned, creating a mosaic of habitats that influences reproductive opportunities for riparian plants and aquatic organisms. Dams along our study streams elevate water levels within the channel and enhance wetlands across the floodplain. The 8 largest wetlands on EFBD Creek total 4.6 hectares over 4 kilometers of stream, while total beaver-generated wetlands on adjacent Alkali Creek total 1.2 hectares over 10 kilometers of stream. All wetlands were delineated during the baseflow period when flooding would not otherwise be occurring.

Intact dams can also enhance channel dynamics by promoting avulsions. In ten years of observations on Odell Creek, two channel avulsions were associated with beaver damming. Each avulsion caused abandonment of a channel segment, resulting in sediment trapping within the segment and initiation of sediment stabilization as vegetation began developing on bars (Supplementary Fig. 1). One of the avulsions created a new wetland adjacent to the redirected channel. We observed additional beaver dams producing overbank flow during seasons and low-flow years when flooding would not otherwise be occurring. Raising local flow stage creates a patchwork of disturbance environments, affecting riparian plant communities through inundation die-offs, and initiating primary succession on bare mineral substrate left behind by receding floodwaters, as documented in other studies18.


In all states of repair, beaver dams induce disturbance that enhances stream dynamics and riparian ecosystem function. Active, intact dams have been recognized as beneficial to restoring incised streams with disconnected floodplains19,20, and for some time have been seen as an important tool in stream restoration21. Our observations provide additional support for the impacts of intact dams. At any point in time, however, active beaver dams affect a small proportion (<5%) of the total channel length on our Montana study streams. Breached or failing beaver dams nonetheless have persistent effects on fluvial processes and can produce desirable restoration outcomes17,22,23. Breaches, where part of the dam remains intact, are common within our study sites, and on other North American streams24. Dams usually breach under high stress during flood discharges24, typically during the snowmelt pulse on the Montana study streams17.

Like large woody debris25,26, the remnants of breached beaver dams perturb water surface slope and add channel roughness, especially where cuttings have sprouted27. Both the decrease in slope and the added roughness promote sediment retention, raising bed elevations near the dam and creating a topographic high that begins to act like a point bar17. In the field of river restoration, a project is usually considered a failure when constructed elements “blow out”28. In the case of beaver dams, however, the remnants of a “blow out” continue to promote habitat diversity beyond the effects of the initial structure. Dam breaching is a critical part of the beaver cycle in streams and acts as an additional disturbance in the riparian corridor, an outcome desired in many restoration projects29.

The addition of small-diameter willow cuttings may act in a similar way to dam remnants, though the effect is less direct. Hydraulic roughness is increased by their accumulation, and as additional cuttings are trapped and roots are established, the effect will increase. Small roughness elements—including larger gravel30, grass, and willow sprouts—can stabilize bar sediments and promote additional sedimentation31. In developing meanders, inner-bank deposition is the predominant process during frequent, small floods32. As the point bar builds outward, aided by roughness from willow cuttings, more flow is directed toward the outer cutbank33 and shear stress is reduced on the inside of the bend.

More broadly, the cycling of beaver activity through the channel network promotes habitat heterogeneity and enhances reproductive opportunities for riparian plants. At any given time, a typical functioning beaver stream encompasses a mosaic of site types, including intact dams, recently breached or abandoned dams, and long-abandoned dams, often interspersed with reaches unsuitable for beaver colonization34. Beaver relocate when resources are depleted at their dam site and move to where preferred food is available35. Some sites can sustain a beaver colony for many years, while some may be occupied for only a brief interval2. Site occupancy also depends on suitable geomorphic characteristics of a particular site, which varies between reaches and catchments16,17,34.

Shifting occupancy by beaver creates ever-changing areas of fresh, moist sediment, including some where vegetation is also recovering following a die-off from inundation27,36. For riparian vegetation, the shifting mosaic of beaver activity can generate variability in ages and species distributions2,6,36 that also results in spatial variations in overbank flow on floodplains. For example, with their closely spaced stems and dense low canopies, willows promote deposition, and offer more protection from erosion in extreme flood flows than cottonwoods27. Riparian communities along streams that lack beaver or large wood are dependent on flood disturbance and channel migration to provide colonization sites on bare sediment37,38, thus shrub and tree distribution is closely tied to flood occurrence. Flood levels must also be fortuitously timed with seed dispersal39. The result is vegetation in cohorts with ages tied to flood events38. In beaver-occupied stream systems, however, beaver produce a steady supply of cuttings and substrates for regeneration (Figs 2, 4), allowing plant recruitment to occur throughout the growing season40. Beaver are most active during the summer months. Declining discharges could strand beaver-cut stems over a range of elevations below crest stage, just as seeds are stranded by falling flows41. The data indicate that moderate to low flows deposit most of the cuttings, coincident with times when beaver are most actively harvesting willow during declining discharge after snowmelt runoff peaks in late May and early June.

Our work also shows that there is some variability in the degree to which beavers disturb a channel that is related, in part, to stream discharge and sinuosity. Frequent breaching, and therefore the ability of breached dams to affect channel morphology, is more common in streams with greater discharge17, while the point bars of more sinuous streams provide sites for establishment of seedlings during frequent bankfull flood events42. Consistent with our findings, the downstream sides of point bars were important sites for establishment of willow vegetative propagules following a large flood that occurred out of phase with willow seed release40. Additionally, lateral accretion of material in migrating bends provides increasing shelter from scour for saplings42,43 Overall, despite beaver impacts being partially dependent on catchment characteristics, the variability in beaver occupancy in both space and time creates riparian patchiness and habitat heterogeneity along the stream corridor17,34,44 with beaver enhancing multiple pathways for willow reproduction (Fig. 6).

Figure 6

Overview of beaver enhancement of willow reproduction through regeneration from the parent plant (asexual reproduction), seed (sexual reproduction) and vegetative propagules (asexual reproduction from plant pieces that have broken off the parent plant). Each process is briefly summarized, basic requirements for the processes (needs) are outlined, and the beaver enhancements are listed.

In addition to reproduction through seeds, which is likely dominant, beaver herbivory and construction of dams and lodges provide a secondary pathway for willow reproduction through the generation of abundant cuttings. Our data show that beaver cuttings are a prolific source of plant propagules to stream networks. Prior work on willow reproduction indicates that floodplain willow cuttings successfully reproduce under a wide range of moisture, sediment and burial regimes—particularly in comparison to non-floodplain willow45—supporting vegetative propagules as a viable reproductive pathway in dynamic floodplain environments46. Genetic data show that sexual reproduction likely dominates in Salicaceae, but the ratio of asexual and sexual reproduction can vary greatly even between different populations of the same species, with variability attributable to local site factors46 – which could include the presence of beaver. Relative to seeds, willow vegetative propagules have a greater reserve of carbohydrates and water within the stem, so are more viable under adverse conditions8,47. Without generation of propagules, willows depend on the large quantities of short-lived, non-dormant seeds produced each year7,48. Generation of seedlings is highly dependent on newly deposited moist sediment, which in the absence of beaver may occupy a very small percentage of the landscape18, and may not be available during the short window of seed viability40. Additionally, seeds of willow appear to have a limited range of dispersal18, whereas vegetative propagules can float downstream for comparatively long distances49, despite trapping by beaver dams and stranding on point bars and other depositional sites along the way. The reproductive strategies of willow along with our findings from Montana streams lead us to hypothesize that the vegetative propagule mode of reproduction should increase in relation to the abundance of beaver in a stream, at least where population size is not strongly limited by food resources, but this remains to be tested.

Our observations of large cutting accumulations on point bars of many other streams in western Montana (e.g., Upper Ruby River, Beaverhead River and tributaries, and Upper Clark Fork River) and across the Rocky Mountain region indicate that fine woody debris accumulations from beaver herbivory are common. In addition, 14C dating of these accumulations in Holocene floodplain sediments along Odell Creek and in the greater Yellowstone region16,50 shows that cuttings are stored over timescales of millennia (Figs 1, 5), showing that an active beaver dam cycle in headwater streams can both affect channel morphodynamics and promote the long-term burial of organic carbon within floodplain sediments. Headwater alluvial streams with actively migrating channels and beaver activity are important locales for carbon storage in mountain rivers, accounting for ~75% of total carbon in the stream network9. Wohl51 estimated that wet beaver meadows can account for 23% of the total carbon stored in the landscape, but this may be a conservative estimate of beaver contributions to carbon storage, as beaver-cut stems in point-bar deposits were not included. Along unconfined meadow reaches of mountain streams where large wood is rare, beaver activity may be the predominant mechanism for carbon storage. The multi-millennial preservation of beaver-chewed wood in point-bar sequences along Odell Creek attests to the importance of beavers in carbon storage well beyond dam sites.

Without beaver, initiation of primary or secondary succession requires flood events or channel migration38. The loss of beaver from stream systems across northern latitudes has meant that much of the research on functioning riparian systems, riparian plant establishment, and the interplay of water and sediment in streams, is grounded in research from sites lacking beaver2. Our work shows, however, that beaver can facilitate multiple pathways for plant colonization (Figs 2, 6) and add to both channel and riparian habitat diversity even after dams are abandoned. Riparian plants reproduce through seed, vegetative propagules and regeneration from damaged plant material, with each pathway augmented by, or predicated on, disturbance. Like other types of disturbance, beaver activity resets succession, allowing pioneer species to gain a foothold. In North American and Eurasia, the full cycle of beaver activity – including dam building, breaching and herbivory – enhances fluvial and floodplain dynamics, and promotes the reproduction and regeneration of riparian plants, not only at active dam sites, but throughout intervening stream reaches as well. Evaluating beaver activity for system-wide effects should be considered in the research and management of fluvial and riparian system dynamics.


Study area

Investigations focused on Odell Creek, with ancillary data on three other streams, East Fork of Blacktail Deer Creek (EFBD Creek), Red Rock Creek and Alkali Creek. All study streams are within the upper headwaters of the Missouri River system in southwestern Montana, USA (Fig. 1). The study reaches on Odell Creek and Red Rock Creek are located within Red Rock Lakes National Wildlife Refuge, and EFBD Creek and Alkali Creek study reaches are located within the Gravelly-Blacktail Wildlife Management Area, managed by Montana Department of Fish, Wildlife and Parks. All study reaches have experienced limited human impact, so provide a valuable reference for studying intact abiotic-biotic interactions. Annual hydrographs across the region are dominated by snowmelt, with peak flows occurring in May and June. The three large streams feature gravel-bed, pool-riffle meandering channels and broad floodplains over most of the study reaches; Alkali Creek shows low-sinuosity meandering in some sections, but is predominantly a plane-bed channel (after ref.52) and has a narrow floodplain in areas lacking beaver (Table 1).

Beaver cutting accumulations

Along Odell Creek we measured the total length of beaver cuttings at 90 sites, within three randomly selected 800 m stream reaches spread over 4 km (Data available at The sites were selected using a stratified random sample based on 3 channel morphologic classes: point bar, straight reach, or cutbank. Within each site, a 0.5 × 0.5 m quadrat was placed at the late summer flow stream edge, as well as at 1 m and 3 m from the active channel, to measure willow cutting accumulation as a function of distance from the channel and relative elevation above low-water stage, yielding a total of 270 quadrats. Within each quadrat we summed the lengths of beaver-generated willow cuttings, as identified by clearly beaver-chewed ends, and counted the number of sprouts growing on the willow cuttings. We also recorded the surface sediment size that covered >50% of the quadrat, from silt through cobble gravel, using the Wentworth grain size classification53. The results for cutting accumulations along Odell Creek showed that as depositional loci, point bars were significantly more important for storing cuttings (Fig. 4 (b)), so we exclusively sampled point bars along three 800 m reaches on Red Rock Creek and five 800 m reaches on EFBD Creek

Along Odell Creek we recorded lengths of beaver generated willow cuttings within the quadrats. Using the program R54, we performed linear mixed effects analyses10 to assess parameters that may influence where and how beaver cuttings are deposited. In each model, the random effect was site name and the fixed effects were the variables of the parameter of interest. To test how each parameter affected mean willow cutting length for a site, we used Chi-squared likelihood ratio tests where the full model, including the parameter of interest, was tested against the null model without the parameter of interest.

Beaver occupancy surveys

Walking surveys, conducted by Montana Fish, Wildlife and Parks and The University of Montana Western, and long-term monitoring on Red Rock and Odell Creeks, along with comparison of sites from National Agriculture Imagery Program (NAIP) imagery were used to document the presence and persistence of individual beaver dams. Data from all methods were used to determine beaver occupancy and abandonment periods.

Subfossil beaver cutting accumulations

Beaver cuttings were collected for radiocarbon dating during stratigraphic investigations of streambank exposures. The beaver cuttings, generated in the stream corridor and quickly buried by progressive point bar sedimentation before decay, are not likely to have been reworked from older deposits, as uncharred woody materials previously preserved in water-saturated sediments are likely to decay rapidly when re-exposed and subjected to an oxygenated environment. The cuttings should thus reflect the age of surrounding sediments, as well as the timing of beaver activity. Rootlets were carefully removed from all wood and charcoal prior to dating. Standard pretreatment procedures55 were used prior to analysis. Samples were 14C dated using accelerator mass spectrometry (AMS)56 at the University of Arizona AMS facility. Three samples were sent to Beta Analytic Incorporated and were dated using radiometric methods requiring larger sample size. Radiocarbon ages were calibrated to calendar years before present (cal yr BP) using CALIB and the IntCal13 calibration curve57. For simplicity of discussion within the paper, individual 14C ages are reported using the median probability of the calibrated age distribution, providing a central point estimate (Fig. 5)58.

Data Availability

The datasets generated and analyzed during the current study are available in the Open Science Framework (OSF) repository at the following address:


  1. 1.

    Stella, J. C. et al. The role of abandoned channels as refugia for sustaining pioneer riparian forest ecosystems. Ecosystems 14, 776–790 (2011).

    Article  Google Scholar 

  2. 2.

    Naiman, R. J., Johnston, C. A. & Kelley, J. C. Alterations of North American streams by beaver. BioScience 38, 753–762 (1988).

    Article  Google Scholar 

  3. 3.

    Westbrook, C. J., Cooper, D. J. & Baker, B. W. Beaver dams and overbank floods influence groundwater–surface water interactions of a Rocky Mountain riparian area. Water Resources Research 42, 1–12, (2006).

    Article  Google Scholar 

  4. 4.

    Puttock, A., Graham, H. A., Cunliffe, A. M., Elliott, M. & Brazier, R. E. Eurasian beaver activity increases water storage, attenuates flow and mitigates diffuse pollution from intensively-managed grasslands. Sci. Total Environ. 576, 430–443 (2017).

    ADS  CAS  Article  Google Scholar 

  5. 5.

    Johnston, C. A. & Naiman, R. J. Aquatic patch creation in relation to beaver population trends. Ecology 71, 1617–1621 (1990).

    Article  Google Scholar 

  6. 6.

    Mouw, J. E. B. et al. Recruitment and successional dynamics diversify the shifting habitat mosaic of an Alaskan floodplain. River Research & Applications 29, 671–685, (2013).

    Article  Google Scholar 

  7. 7.

    Cottrell, T. R. Willow colonization of Rocky-Mountain mires. Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere 25, 215–222, (1995).

    Article  Google Scholar 

  8. 8.

    Krasny, M. E., Vogt, K. A. & Zasada, J. C. Establishment of four Salicaceae species on river bars in interior Alaska. Holarctic Ecology 11, 210–219 (1988).

    Google Scholar 

  9. 9.

    Wohl, E., Dwire, K., Sutfin, N., Polvi, L. & Bazan, R. Mechanisms of carbon storage in mountainous headwater rivers. Nature Communications 3, 1–8 (2012).

    Article  Google Scholar 

  10. 10.

    Bates, D., Maechler, M., Bolker, B. & Walker, S. Fitting linear mixed-effects models using lme4. Journal of Statistical Software 67, 1–48, (2015).

    Article  Google Scholar 

  11. 11.

    Dietrich, W. E. & Smith, J. D. Bed-load transport in a river meander. Water Resources Research 20, 1355–1380, (1984).

    ADS  Article  Google Scholar 

  12. 12.

    Woods, S. W. & Cooper, D. J. Hydrologic Factors Affecting Initial Willow Seedling Establishment along a Subalpine Stream, Colorado, USA. Arctic, Antarctic, and Alpine Research 37, 636–643 (2005).

    Article  Google Scholar 

  13. 13.

    Woelfle‐Erskine, C., Wilcox, A. C. & Moore, J. N. Combining historical and process perspectives to infer ranges of geomorphic variability and inform river restoration in a wandering gravel‐bed river. Earth Surface Processes and Landforms 37, 1302–1312 (2012).

    ADS  Article  Google Scholar 

  14. 14.

    Millington, C. E. & Sear, D. A. Impacts of river restoration on small-wood dynamics in a low-gradient headwater stream. Earth Surface Processes and Landforms 32, 1204–1218, (2007).

    ADS  Article  Google Scholar 

  15. 15.

    Nanson, G. C. Point bar and floodplain formation of the meandering Beatton River, northeastern British Columbia, Canada. Sedimentology 27, 3–29 (1980).

    ADS  Article  Google Scholar 

  16. 16.

    Persico, L. & Meyer, G. Holocene beaver damming, fluvial geomorphology, and climate in Yellowstone National Park, Wyoming. Quaternary Research 71, 340–353, (2009).

    ADS  Article  Google Scholar 

  17. 17.

    Levine, R. & Meyer, G. A. Beaver dams and channel sediment dynamics on Odell Creek, Centennial Valley. Montana, USA Geomorphology 205, 51–64, (2014).

    ADS  Article  Google Scholar 

  18. 18.

    Gage, E. A. & Cooper, D. J. Patterns of willow seed dispersal, seed entrapment, and seedling establishment in a heavily browsed montane riparian ecosystem. Canadian Journal of Botany 83, 678–687, (2005).

    Article  Google Scholar 

  19. 19.

    Beechie, T. J., Pollock, M. M. & Baker, S. Channel incision, evolution and potential recovery in the Walla Walla and Tucannon River basins, northwestern USA. Earth Surface Processes and Landforms 33, 784–800, (2008).

    ADS  Article  Google Scholar 

  20. 20.

    Pollock, M. M., Beechie, T. J. & Jordan, C. E. Geomorphic changes upstream of beaver dams in Bridge Creek, an incised stream channel in the interior Columbia River basin, eastern Oregon. Earth Surface Process and Landforms 32, 1174–1185 (2007).

    ADS  Article  Google Scholar 

  21. 21.

    Apple, L. L. Riparian habitat restoration and beavers. Riparian ecosystems and their management: reconciling conflicting uses. US Forest Service General Technical Report RM. 120, 489–490 (1985).

    Google Scholar 

  22. 22.

    Bouwes, N. et al. Ecosystem experiment reveals benefits of natural and simulated beaver dams to a threatened population of steelhead (Oncorhynchus mykiss). Scientific reports 6, 28581 (2016).

    ADS  CAS  Article  Google Scholar 

  23. 23.

    Pollock, M. M. et al. Using Beaver Dams to Restore Incised Stream Ecosystems. BioScience 64, 279–290, (2014).

    Article  Google Scholar 

  24. 24.

    Demmer, R. & Beschta, R. L. Recent history (1988-2004) of beaver dams along Bridge Creek in Central Oregon. Northwest Science 82, 309–318 (2008).

    Article  Google Scholar 

  25. 25.

    Buffington, J. M. & Montgomery, D. R. A systematic analysis of eight decades of incipient motion studies, with special reference to gravel-bedded rivers. Water Resources Research 33, 1993–2029 (1997).

    ADS  Article  Google Scholar 

  26. 26.

    Brummer, C. J., Abbe, T. B., Sampson, J. R. & Montgomery, D. R. Influence of vertical channel change associated with wood accumulations on delineating channel migration zones, Washington, USA. Geomorphology 80, 295–309 (2006).

    ADS  Article  Google Scholar 

  27. 27.

    Smith, J. D. In Plant disturbance ecology: the process and the response. (eds Edward A. Johnson & Kiyoko Miyanishi) 603–671 (Elsevier Academic Press, 2007).

  28. 28.

    Simon, A. et al. Critical evaluation of how the Rosgen classification and associated “Natural Channel Design” methods fail to integrate and quantify fluvial processes and channel response. JAWRA Journal of the American Water Resources Association 43, 1117–1131, (2007).

    ADS  Article  Google Scholar 

  29. 29.

    Richards, K., Brasington, J. & Hughes, F. Geomorphic dynamics of floodplains: ecological implications and a potential modelling strategy. Freshwater Biology 47, 559–579, (2002).

    Article  Google Scholar 

  30. 30.

    Fryirs, K. A., Brierley, G. J., Preston, N. J. & Kasai, M. Buffers, barriers and blankets: The (dis)connectivity of catchment-scale sediment cascades. Catena 70, 49–67, (2007).

    Article  Google Scholar 

  31. 31.

    Parker, G. et al. A new framework for modeling the migration of meandering rivers. Earth Surf. Process. Landf. 36, 70–86, (2011).

    ADS  Article  Google Scholar 

  32. 32.

    Pizzuto, J. E. Channel adjustments to changing discharges, Powder River, Montana. Geological Society of America Bulletin 106, 1494–1501 (1994).

    ADS  Article  Google Scholar 

  33. 33.

    Dietrich, W. E. & Smith, J. D. Influence of the point-bar on flow through curved channels. Water Resources Research 19, 1173–1192, (1983).

    ADS  Article  Google Scholar 

  34. 34.

    Macfarlane, W. W. et al. Modeling the capacity of riverscapes to support beaver dams. Geomorphology 277, 72–99 (2017).

    ADS  Article  Google Scholar 

  35. 35.

    Donkor, N. T. & Fryxell, J. M. Impact of beaver foraging on structure of lowland boreal forests of Algonquin Provincial Park, Ontario. Forest Ecology and Management 118, 83–92, (1999).

    Article  Google Scholar 

  36. 36.

    Amlin, N. A. & Rood, S. B. Inundation tolerances of riparian willows and cottonwoods. JAWRA Journal of the American Water Resources Association 37, 1709–1720, (2001).

    ADS  Article  Google Scholar 

  37. 37.

    Johnson, W. C. et al. Forty years of vegetation change on the Missouri River floodplain. BioScience 62, 123–135, (2012).

    Article  Google Scholar 

  38. 38.

    Merritt, D. M. & Cooper, D. J. Riparian vegetation and channel change in response to river regulation: a comparative study of regulated and unregulated streams in the Green River Basin, USA. Regulated Rivers: Research & Management 16, 543–564 (2000).

    Article  Google Scholar 

  39. 39.

    Rood, S. B. et al. Managing river flows to restore floodplain forests. Frontiers in Ecology and the Environment 3, 193–201,[0193:mrftrf];2 (2005).

    Article  Google Scholar 

  40. 40.

    Asaeda, T., Gomes, P. I. A., Sakamoto, K. & Rashid, M. H. Tree colonization trends on a sediment bar after a major flood. River Research and Applications 27, 976–984, (2011).

    Article  Google Scholar 

  41. 41.

    Merritt, D. M. & Wohl, E. E. Processes governing hydrochory along rivers: hydraulics, hydrology, and dispersal phenology. Ecological Applications 12, 1071–1087,[1071:PGHARH]2.0.CO;2 (2002).

    Article  Google Scholar 

  42. 42.

    Scott, M. L., Friedman, J. M. & Auble, G. T. Fluvial process and the establishment of bottomland trees. Geomorphology 14, 327–339, (1996).

    ADS  Article  Google Scholar 

  43. 43.

    Corenblit, D. et al. The biogeomorphological life cycle of poplars during the fluvial biogeomorphological succession: a special focus on Populus nigra L. Earth Surface Processes and Landforms 39, 546–563, (2014).

    ADS  Article  Google Scholar 

  44. 44.

    Polvi, L. E. & Wohl, E. The beaver meadow complex revisited - the role of beavers in post-glacial floodplain development. Earth Surface Process and Landforms 37, 332–346, (2012).

    ADS  Article  Google Scholar 

  45. 45.

    Radtke, A., Mosner, E. & Leyer, I. Vegetative reproduction capacities of floodplain willows–cutting response to competition and biomass loss. Plant Biology 14, 257–264 (2012).

    CAS  Article  Google Scholar 

  46. 46.

    Karrenberg, S., Edwards, P. J. & Kollmann, J. The life history of Salicaceae living in the active zone of floodplains. Freshwater Biology 47, 733–748, (2002).

    Article  Google Scholar 

  47. 47.

    Thomas, L. K., Toelle, L., Ziegenhagen, B. & Leyer, I. Are vegetative reproduction capacities the cause of widespread invasion of Eurasian Salicaceae in Patagonian river landscapes? Plos One 7, e50652, (2012).

    ADS  CAS  Article  PubMed  PubMed Central  Google Scholar 

  48. 48.

    Densmore, R. & Zasada, J. Seed dispersal and dormancy patterns in northern willows: ecological and evolutionary significance. Canadian Journal of Botany-Revue Canadienne De Botanique 61, 3207–3216 (1983).

    Google Scholar 

  49. 49.

    Boedeltje, G., Bakker, J. P., Ten Brinke, A., Van Groenendael, J. M. & Soesbergen, M. Dispersal phenology of hydrochorous plants in relation to discharge, seed release time and buoyancy of seeds: the flood pulse concept supported. Journal of Ecology 92, 786–796, (2004).

    Article  Google Scholar 

  50. 50.

    Persico, L. & Meyer, G. Natural and historical variability in fluvial processes, beaver activity, and climate in the Greater Yellowstone Ecosystem. Earth Surface Processes and Landforms 38, 728–750, (2013).

    ADS  Article  Google Scholar 

  51. 51.

    Wohl, E. Landscape-scale carbon storage associated with beaver dams. Geophysical Research Letters 40, 3631–3636, (2013).

    ADS  CAS  Article  Google Scholar 

  52. 52.

    Montgomery, D. R. & Buffington, J. M. Channel-reach morphology in mountain drainage basins. Geological Society of American 109, 595–611 (1997).

    ADS  Google Scholar 

  53. 53.

    Wentworth, C. K. A scale of grade and class terms for clastic sediments. The Journal of Geology 30, 377–392 (1922).

    ADS  Article  Google Scholar 

  54. 54.

    R: A Language and Environment for Statistical Computing (Vienna, Austria, 2015).

  55. 55.

    Brock, F., Higham, T., Ditchfield, P. & Bronk Ramsey, C. Current pretreatment methods for AMS radiocarbon dating at the Oxford Radiocarbon Accelerator Unit (ORAU). Radiocarbon 52, 103–112 (2010).

    CAS  Article  Google Scholar 

  56. 56.

    Dickin, A. P. Radiogenic Isotope Geology. (Cambridge University Press, 2005).

  57. 57.

    Reimer, P. J. et al. IntCal13 and Marine13 radiocarbon age calibration curves 0–50,000 years cal BP. Radiocarbon 55, 1869–1887 (2013).

    CAS  Article  Google Scholar 

  58. 58.

    Telford, R. J., Heegaard, E. & Birks, H. J. B. The intercept is a poor estimate of a calibrated radiocarbon age. The Holocene 14, 296–298, (2004).

    ADS  Article  Google Scholar 

  59. 59.

    McCarthy, P. M., Sando, R., Sando, S. K. & Dutton, D. M. Methods for estimating streamflow characteristics at ungaged sites in western Montana based on data through water year 2009: Chapter G in Montana StreamStats. Report No. 2015-5019G, 32 (Reston, VA, 2016).

Download references


We would like to thank the National Science Foundation (grant EAR-0844293 to Meyer), the Earth and Planetary Science Department at the University of New Mexico, Red Rock Lakes National Wildlife Refuge, the University of Montana Western (UMW) and The Nature Conservancy of Montana for major project funding and support. Lidar data was provided by the National Center for Airborne Laser Swath Mapping (grant to Levine under EAR-1043051). Field and lab assistance was provided by Felisha Yazzie, Michelle Gauthier, Cory Burkett, Waheulla Barney, Kyle Cutting, Andrea Kristof, Jeff Warren and UMW hydrology students. Land access and beaver survey data were provided by Montana Fish, Wildlife and Parks. Comments by Leslie McFadden improved an earlier draft of this manuscript; two anonymous reviewers provided insights that further improved the paper.

Author information




R.L. conceived of, designed and conducted field work for the study. G.M. provided critical feedback to improve study design and supervised the work. R.L. wrote the initial draft of the manuscript. Both authors discussed results, provided critical feedback and contributed to the final version of the manuscript.

Corresponding author

Correspondence to Rebekah Levine.

Ethics declarations

Competing Interests

The authors declare no competing interests.

Additional information

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Levine, R., Meyer, G.A. Beaver-generated disturbance extends beyond active dam sites to enhance stream morphodynamics and riparian plant recruitment. Sci Rep 9, 8124 (2019).

Download citation


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing