CLBMON-12 Arrow Lakes Reservoir Monitoring of Revegetation Efforts and Vegetation Composition Analysis

Author: M.T. Miller, P. Gibeau, V.C. Hawkes, Okanagan Nation Alliance, B.C. Westbank, L.G.L.

This study assesses the effectiveness of revegetation treatments applied in the drawdown zone of Arrow Lakes Reservoir under CLBWORKS-2 between 2009-2011. The primary objective is to assess the short-term effectiveness of the revegetation program at expanding the quality and quantity of vegetation in the drawdown zone for ecological and social benefits. This monitoring program also assesses the intra-community response of existing vegetation in the drawdown zone to the Water Use Planning operating regime at the site level.

*For complete summary of key results and relevant management questions, please refer to PP.III to X of the document.
Monitoring of Revegetated Areas
MQ3. What environmental conditions, including the current operating regime (i.e. timing, frequency, duration and depth of inundation), may limit or improve the restoration and expansion of vegetation communities in the drawdown zone?
Summary Findings
Revegetation performance appears to be correlated to a range of environmental factors including inundation, infertile or unstable substrates, wave action and erosion, sedimentation, and soil moisture deficits.
All aspects of the operating regime have the potential to limit or improve the restoration and expansion of vegetation communities. Timing of inundation determines the ability of restored vegetation to set roots, grow, and reproduce within the annual cycle. Frequency of inundation can affect establishment rates, especially of woody species at upper elevations. Duration and depth of inundation determine the levels of anoxia that plants must endure and the degree of seasonal exposure to wave action, erosion, sedimentation, and woody debris.
Sources of Uncertainty/ Limitations
Insufficient treatment replications (both spatially and temporally) limit our ability to directly correlate revegetation effectiveness with different operational components (i.e., timing, frequency, duration and depth of inundation), and to separate these effects from other, non-operational effects.
Comments
Physical works projects aimed at establishing vegetation in the Arrow Lake Reservoir drawdown zone should strive to ensure that adequate experimental replication (including spatial and temporal replication) is incorporated as an intrinsic component of any future revegetation prescriptions.
MQ6. Is there an opportunity to modify operations to more effectively maintain revegetated communities at the landscape and site level in the future?
Summary Findings
In theory, opportunities exist for modifying operations to help restoration goals, but this idea has not been adequately tested. Soft constraints were operational targets. The reservoir was not operated to target specific soft constraints, although the general operation under the water use plan allowed for variation where the soft constraint for vegetation was partially met. From 2008 to 2017, the soft constraint target for vegetation (‰¤ 434 m ASL between April and October) was met 46% of the time.
Experience with the revegetation program to date suggests that operations will be most effective at maintaining revegetated communities to the extent they are employed to limit not just the depth but also the duration of inundation during the summer and early fall growing season.
Sources of Uncertainty/ Limitations
As noted above (MQ 3), insufficient replication of alternative operational regimes, and of the 2008-2011 revegetation treatments across elevation bands, habitat types, and years, precludes testing of hypotheses around revegetation efficacy as it relates to operational (reservoir-related) and non-operational (environmental) factors.
Monitoring of Existing Vegetation
MQ3. How does the current operating regime affect the within-community quality and quantity (i.e., species cover, abundance, biomass, diversity and distribution within existing communities) of existing vegetation?
Summary Findings
As noted above, operation of the reservoir has resulted in identifiable vegetation zonation patterns within the drawdown zone between 434 m and 440 m that are correlated to varying degree with elevation, reservoir operations, and topo-edaphic features. To the best of our knowledge, the current operating regime has succeeded in maintaining these basic zonation patterns over time.
Sites that undergo prolonged inundation (>100 days) tend to have lower vegetation cover, height, biomass, and abundance than sites with a longer exposure period (Enns and Overholt 2013a). However, coarse-textured soils with poor water-holding capacity at high elevation can also show reductions in these traits, presumably due to drought stress. Some species and communities likely benefit from seasonal inundation, though not from scouring or from wave action on exposed high energy sites. In the fall of 2015, after an unusually low summer reservoir maximum of 435.5 m, vegetation growing above the summer flood line was more vigorous and appeared, in general, to benefit from the yearlong release from inundation. However, at some sedge-dominated sites a band of greener, lusher foliage was observed just below the flood line and several species were undergoing a late-season flowering pulse or were dispersing seed. The increase in available soil moisture following brief inundation likely allowed some species to extend their growing season into the fall, suggesting that in some cases the short-term benefits of brief inundation may exceed or at least equal those accruing from non-inundation (Miller et al. 2016).
With the exception of the Beach (BE) VCT, where diversity may have increased, and the PC-reed canarygrass VCT, where diversity may have decreased, there were few notable directional changes in diversity between 2010 and 2016, implying that the composition of VCTs has remained relatively stable over time (Miller et al. 2018). However, there have been small but statistically significant declines in overall per cent plant cover at all elevations, and in the cover of some plant guilds (forbs, sedges and sedge allies, pteridophytes), between 2010 and 2016. Per cent cover of grasses has not changed, while that of shrubs may have increased slightly. At mid elevation, forb cover appears to have decreased relative to cover of graminoids, pteridophytes, and shrubs. At low elevation, the proportional cover of grasses versus that of sedges and sedge-like plants has fluctuated over time. Whereas the two groups had similar coverage on average in 2013, grasses were more than twice as abundant (on average) as sedges and their allies in 2016.
Since 2012, the reservoir has experienced an incremental increase in seasonal growing degree days (GDDs) at low elevation that appears to coincide with the declining trend in the covers of forbs, sedges, and pteridophytes over the same period. At mid elevation, where total cumulative GDDs declined from 2010 until the middle of the monitoring period before increasing again, cover of grasses has tended to follow a slightly divergent pattern, with highest average covers recorded in 2012 and lower covers thereafter. The implication is that, with the probable exception of shrubs, decreases in late summer inundation do not necessarily translate into an increase in plant density or abundance at the local scale.
Sources of Uncertainty/ Limitations
– Lack of a formal study (experimental) control, which is necessary to separate operational effects from other environmental effects (e.g., annual climatic variation).
– Lack of historical baseline information on the conditions that pertained prior to introduction of soft constraints.
– Thedurationofthismonitoringprogrammaynothavebeenlongenoughtofullyassessthelong-termeffectsofthecurrentoperatingregime on the structure and composition of existing vegetation communities.
Comments
Soft constraints were operational targets. The reservoir was not operated to target specific soft constraints; however, the general operation under the water use plan allowed for variation where the soft constraint for vegetation was partially met. From 2008 to 2016, the soft constraint target for vegetation (‰¤ 434 m ASL between April and October) was met 48% of the time.
The variable annual reservoir operations that have prevailed since the start of the study in 2007, in combination with the biannual sampling regime, limit our ability to test specific hypotheses around impacts stemming from alterations to the frequency, timing, depth, and duration of inundation.
MQ4. Is there a shift in community structure (e.g., species dominance) or a potential loss of existing vegetated communities that is attributable to environmental conditions, including the current operating regime (i.e., timing, frequency, duration and depth of inundation)?
Summary Findings
Temporal, localized fluctuations in cover (Enns and Overholt 2013a; Miller et al. 2015; 2018), together with a pattern of directional declines observed between 2010 and 2016 for some plant groups such as forbs and sedges, and increases for some groups such as shrubs, raises reasonable doubts about the effectiveness of the current operating regime in maintaining the vegetation status quo at the site level. The drawdown zone supports a vegetation assemblage that is adapted to, and may even depend on, a variable regime of seasonal flooding as part of annual moisture requirements. Thus, we can expect to see directional changes in vegetation cover and composition in response to any consistent, directional changes in the timing, depth, frequency, and duration of inundation. For example, recent monitoring suggests that successive years of above-average summer exposure could lead to a more shrub-dominated system supporting lower overall covers of herbaceous groups such as forbs and sedges. While there is currently no compelling evidence to indicate that the Water Use Plan operating regime is failing to maintain overall vegetation structure and composition of existing vegetation communities in the drawdown zone, existing vegetation will likely only be maintained in its present state if the historical pattern of variability in hydroperiod is maintained (Miller et al. 2018).
Sources of Uncertainty/ Limitations
The duration of this monitoring program may not have been long enough to fully assess the long-term effects of the current operating regime on the structure and composition of existing vegetation communities.
MQ5. What are the species- specific survival rates under soft constraints operating regime (i.e., what are the tolerances of existing plant species to inundation)?
Summary Findings
Enns and Enns (2012) reviewed the flood tolerances of a range of common, dominant drawdown zone species. Mortality of weedy annuals, seedlings, and herbaceous perennials was common after extended inundation, but replacements also took place. Invasions from upslope occurred, were lost, and recurred. Willows declined in Revelstoke Reach at mid elevations as a result of populations expanding into these sites and subsequently being inundated for longer than their usual tolerance would allow. Black cottonwood is also largely restricted to upper elevations of the drawdown zone where inundation duration is reduced. Horsetails, along with well-established tussocks of sedges (e.g., Kellogg€™s sedge, Columbia sedge) and other graminoids (e.g., little meadow-foxtail, thread rush, Canada bluegrass, reed canarygrass) are highly robust to episodes of deep and prolonged inundation, but are vulnerable to erosive forces, wave action, and sediment deposition on exposed aspects (Enns and Overholt 2013a, Miller et al. 2018).
Sources of Uncertainty/ Limitations
The original study design (Enns et al. 2007) did not include a replicated, controlled approach for directly estimating species-specific mortality, or for distinguishing inundation impacts on survivorship from other environmental impacts (e.g., herbivory, drought). Thus, species-specific survivorship rates, as these relate to inundation, can only be inferred indirectly based on observed changes in plant composition over time within repeat monitoring plots. The potential for prolonged (> 1 year) time lags in plant responses to changing hydro regimes also limits our ability test hypotheses around operational impacts on plant survival.
This MQ will be more thoroughly addressed under CLBMON-35 (program in progress) for a select group of key species.
Comments
Soft constraints were operational targets. The reservoir was not operated to target specific soft constraints; however, the general operation under the water use plan allowed for variation where the soft constraint for vegetation was partially met. From 2008 to 2016, the soft constraint target for vegetation (‰¤ 434 m ASL between April and October) was met 48% of the time.
MQ6. What recommendations can be made to more effectively maintain existing vegetation at the site level in the future?
Summary Findings
The drawdown zone of Arrow Lakes Reservoir supports a vegetation assemblage that is adapted to, and may even depend on, a variable regime of seasonal flooding as part of annual moisture requirements. The best way to ensure that the soft constraints operating regime continues to maintain the existing vegetation status quo is to maintain a similar level of variability in hydroperiod to that which has prevailed historically.
if the objective is to enhance existing vegetation types, our models suggest that both cover and structural diversity at all elevations can be maximized in the following way: (i) by delaying inundation for as long as possible in the spring (preferably until after June), to allow time for germination, establishment, and the completion of reproductive cycles; (ii) by allowing for sufficient June/July inundation at low and mid elevations (434-438 m ASL) to reduce summer drought stress for inundation-adapted species; and (iii) by minimizing (but not eliminating) the depth and duration of inundation at high elevations (>438 m ASL), to maintain herbaceous cover while facilitating woody shrub establishment and growth.
Sources of Uncertainty/ Limitations
The variable annual reservoir operations that have prevailed since the start of the study in 2007, in combination with the biannual sampling regime, limit our ability to predict specific impacts stemming from alterations to the frequency, timing, depth, and duration of inundation. At present we can only offer hypotheses, based on the best available data, around the potential long-term outcomes of different hydroperiod scenarios.
Comments
It may be possible to implement physical works (PW) to either protect or create habitats in the drawdown zone, which could lead to the maintenance of vegetation communities. For example, in Kinbasket Reservoir, elevated mounds and windrows have been constructed at some sites to increase topographic heterogeneity. PW at some sites (e.g., Burton Creek) are currently under consideration. These efforts are localized, small-scale projects that will not result in widespread benefits to drawdown zone vegetation.

*For complete summary of key results and relevant management questions, please refer to PP.III to X of the document.
Monitoring of Revegetated Areas
MQ3. What environmental conditions, including the current operating regime (i.e. timing, frequency, duration and depth of inundation), may limit or improve the restoration and expansion of vegetation communities in the drawdown zone?
Summary Findings
Revegetation performance appears to be correlated to a range of environmental factors including inundation, infertile or unstable substrates, wave action and erosion, sedimentation, and soil moisture deficits.
All aspects of the operating regime have the potential to limit or improve the restoration and expansion of vegetation communities. Timing of inundation determines the ability of restored vegetation to set roots, grow, and reproduce within the annual cycle. Frequency of inundation can affect establishment rates, especially of woody species at upper elevations. Duration and depth of inundation determine the levels of anoxia that plants must endure and the degree of seasonal exposure to wave action, erosion, sedimentation, and woody debris.
Sources of Uncertainty/ Limitations
Insufficient treatment replications (both spatially and temporally) limit our ability to directly correlate revegetation effectiveness with different operational components (i.e., timing, frequency, duration and depth of inundation), and to separate these effects from other, non-operational effects.
Comments
Physical works projects aimed at establishing vegetation in the Arrow Lake Reservoir drawdown zone should strive to ensure that adequate experimental replication (including spatial and temporal replication) is incorporated as an intrinsic component of any future revegetation prescriptions.
MQ6. Is there an opportunity to modify operations to more effectively maintain revegetated communities at the landscape and site level in the future?
Summary Findings
In theory, opportunities exist for modifying operations to help restoration goals, but this idea has not been adequately tested. Soft constraints were operational targets. The reservoir was not operated to target specific soft constraints, although the general operation under the water use plan allowed for variation where the soft constraint for vegetation was partially met. From 2008 to 2017, the soft constraint target for vegetation (‰¤ 434 m ASL between April and October) was met 46% of the time.
Experience with the revegetation program to date suggests that operations will be most effective at maintaining revegetated communities to the extent they are employed to limit not just the depth but also the duration of inundation during the summer and early fall growing season.
Sources of Uncertainty/ Limitations
As noted above (MQ 3), insufficient replication of alternative operational regimes, and of the 2008-2011 revegetation treatments across elevation bands, habitat types, and years, precludes testing of hypotheses around revegetation efficacy as it relates to operational (reservoir-related) and non-operational (environmental) factors.
Monitoring of Existing Vegetation
MQ3. How does the current operating regime affect the within-community quality and quantity (i.e., species cover, abundance, biomass, diversity and distribution within existing communities) of existing vegetation?
Summary Findings
As noted above, operation of the reservoir has resulted in identifiable vegetation zonation patterns within the drawdown zone between 434 m and 440 m that are correlated to varying degree with elevation, reservoir operations, and topo-edaphic features. To the best of our knowledge, the current operating regime has succeeded in maintaining these basic zonation patterns over time.
Sites that undergo prolonged inundation (>100 days) tend to have lower vegetation cover, height, biomass, and abundance than sites with a longer exposure period (Enns and Overholt 2013a). However, coarse-textured soils with poor water-holding capacity at high elevation can also show reductions in these traits, presumably due to drought stress. Some species and communities likely benefit from seasonal inundation, though not from scouring or from wave action on exposed high energy sites. In the fall of 2015, after an unusually low summer reservoir maximum of 435.5 m, vegetation growing above the summer flood line was more vigorous and appeared, in general, to benefit from the yearlong release from inundation. However, at some sedge-dominated sites a band of greener, lusher foliage was observed just below the flood line and several species were undergoing a late-season flowering pulse or were dispersing seed. The increase in available soil moisture following brief inundation likely allowed some species to extend their growing season into the fall, suggesting that in some cases the short-term benefits of brief inundation may exceed or at least equal those accruing from non-inundation (Miller et al. 2016).
With the exception of the Beach (BE) VCT, where diversity may have increased, and the PC-reed canarygrass VCT, where diversity may have decreased, there were few notable directional changes in diversity between 2010 and 2016, implying that the composition of VCTs has remained relatively stable over time (Miller et al. 2018). However, there have been small but statistically significant declines in overall per cent plant cover at all elevations, and in the cover of some plant guilds (forbs, sedges and sedge allies, pteridophytes), between 2010 and 2016. Per cent cover of grasses has not changed, while that of shrubs may have increased slightly. At mid elevation, forb cover appears to have decreased relative to cover of graminoids, pteridophytes, and shrubs. At low elevation, the proportional cover of grasses versus that of sedges and sedge-like plants has fluctuated over time. Whereas the two groups had similar coverage on average in 2013, grasses were more than twice as abundant (on average) as sedges and their allies in 2016.
Since 2012, the reservoir has experienced an incremental increase in seasonal growing degree days (GDDs) at low elevation that appears to coincide with the declining trend in the covers of forbs, sedges, and pteridophytes over the same period. At mid elevation, where total cumulative GDDs declined from 2010 until the middle of the monitoring period before increasing again, cover of grasses has tended to follow a slightly divergent pattern, with highest average covers recorded in 2012 and lower covers thereafter. The implication is that, with the probable exception of shrubs, decreases in late summer inundation do not necessarily translate into an increase in plant density or abundance at the local scale.
Sources of Uncertainty/ Limitations
– Lack of a formal study (experimental) control, which is necessary to separate operational effects from other environmental effects (e.g., annual climatic variation).
– Lack of historical baseline information on the conditions that pertained prior to introduction of soft constraints.
– Thedurationofthismonitoringprogrammaynothavebeenlongenoughtofullyassessthelong-termeffectsofthecurrentoperatingregime on the structure and composition of existing vegetation communities.
Comments
Soft constraints were operational targets. The reservoir was not operated to target specific soft constraints; however, the general operation under the water use plan allowed for variation where the soft constraint for vegetation was partially met. From 2008 to 2016, the soft constraint target for vegetation (‰¤ 434 m ASL between April and October) was met 48% of the time.
The variable annual reservoir operations that have prevailed since the start of the study in 2007, in combination with the biannual sampling regime, limit our ability to test specific hypotheses around impacts stemming from alterations to the frequency, timing, depth, and duration of inundation.
MQ4. Is there a shift in community structure (e.g., species dominance) or a potential loss of existing vegetated communities that is attributable to environmental conditions, including the current operating regime (i.e., timing, frequency, duration and depth of inundation)?
Summary Findings
Temporal, localized fluctuations in cover (Enns and Overholt 2013a; Miller et al. 2015; 2018), together with a pattern of directional declines observed between 2010 and 2016 for some plant groups such as forbs and sedges, and increases for some groups such as shrubs, raises reasonable doubts about the effectiveness of the current operating regime in maintaining the vegetation status quo at the site level. The drawdown zone supports a vegetation assemblage that is adapted to, and may even depend on, a variable regime of seasonal flooding as part of annual moisture requirements. Thus, we can expect to see directional changes in vegetation cover and composition in response to any consistent, directional changes in the timing, depth, frequency, and duration of inundation. For example, recent monitoring suggests that successive years of above-average summer exposure could lead to a more shrub-dominated system supporting lower overall covers of herbaceous groups such as forbs and sedges. While there is currently no compelling evidence to indicate that the Water Use Plan operating regime is failing to maintain overall vegetation structure and composition of existing vegetation communities in the drawdown zone, existing vegetation will likely only be maintained in its present state if the historical pattern of variability in hydroperiod is maintained (Miller et al. 2018).
Sources of Uncertainty/ Limitations
The duration of this monitoring program may not have been long enough to fully assess the long-term effects of the current operating regime on the structure and composition of existing vegetation communities.
MQ5. What are the species- specific survival rates under soft constraints operating regime (i.e., what are the tolerances of existing plant species to inundation)?
Summary Findings
Enns and Enns (2012) reviewed the flood tolerances of a range of common, dominant drawdown zone species. Mortality of weedy annuals, seedlings, and herbaceous perennials was common after extended inundation, but replacements also took place. Invasions from upslope occurred, were lost, and recurred. Willows declined in Revelstoke Reach at mid elevations as a result of populations expanding into these sites and subsequently being inundated for longer than their usual tolerance would allow. Black cottonwood is also largely restricted to upper elevations of the drawdown zone where inundation duration is reduced. Horsetails, along with well-established tussocks of sedges (e.g., Kellogg€™s sedge, Columbia sedge) and other graminoids (e.g., little meadow-foxtail, thread rush, Canada bluegrass, reed canarygrass) are highly robust to episodes of deep and prolonged inundation, but are vulnerable to erosive forces, wave action, and sediment deposition on exposed aspects (Enns and Overholt 2013a, Miller et al. 2018).
Sources of Uncertainty/ Limitations
The original study design (Enns et al. 2007) did not include a replicated, controlled approach for directly estimating species-specific mortality, or for distinguishing inundation impacts on survivorship from other environmental impacts (e.g., herbivory, drought). Thus, species-specific survivorship rates, as these relate to inundation, can only be inferred indirectly based on observed changes in plant composition over time within repeat monitoring plots. The potential for prolonged (> 1 year) time lags in plant responses to changing hydro regimes also limits our ability test hypotheses around operational impacts on plant survival.
This MQ will be more thoroughly addressed under CLBMON-35 (program in progress) for a select group of key species.
Comments
Soft constraints were operational targets. The reservoir was not operated to target specific soft constraints; however, the general operation under the water use plan allowed for variation where the soft constraint for vegetation was partially met. From 2008 to 2016, the soft constraint target for vegetation (‰¤ 434 m ASL between April and October) was met 48% of the time.
MQ6. What recommendations can be made to more effectively maintain existing vegetation at the site level in the future?
Summary Findings
The drawdown zone of Arrow Lakes Reservoir supports a vegetation assemblage that is adapted to, and may even depend on, a variable regime of seasonal flooding as part of annual moisture requirements. The best way to ensure that the soft constraints operating regime continues to maintain the existing vegetation status quo is to maintain a similar level of variability in hydroperiod to that which has prevailed historically.
if the objective is to enhance existing vegetation types, our models suggest that both cover and structural diversity at all elevations can be maximized in the following way: (i) by delaying inundation for as long as possible in the spring (preferably until after June), to allow time for germination, establishment, and the completion of reproductive cycles; (ii) by allowing for sufficient June/July inundation at low and mid elevations (434-438 m ASL) to reduce summer drought stress for inundation-adapted species; and (iii) by minimizing (but not eliminating) the depth and duration of inundation at high elevations (>438 m ASL), to maintain herbaceous cover while facilitating woody shrub establishment and growth.
Sources of Uncertainty/ Limitations
The variable annual reservoir operations that have prevailed since the start of the study in 2007, in combination with the biannual sampling regime, limit our ability to predict specific impacts stemming from alterations to the frequency, timing, depth, and duration of inundation. At present we can only offer hypotheses, based on the best available data, around the potential long-term outcomes of different hydroperiod scenarios.
Comments
It may be possible to implement physical works (PW) to either protect or create habitats in the drawdown zone, which could lead to the maintenance of vegetation communities. For example, in Kinbasket Reservoir, elevated mounds and windrows have been constructed at some sites to increase topographic heterogeneity. PW at some sites (e.g., Burton Creek) are currently under consideration. These efforts are localized, small-scale projects that will not result in widespread benefits to drawdown zone vegetation.