Step Pool Research Articles

Stream modifications to enhance system connectivity for fish habitat compensation: a case study in the Barrenlands region of Canada

This study examines stream modification efforts to increase the productive capacity of an isolated system of three small lakes in the Barrenlands region of Arctic Canada by enhancing system connectivity. The lakes’ outlet streams were modified to create conditions favourable for fish passage and thereby promote migration among the lakes and the large lake into which they drain. Gabion step pools (in two streams) and a nature-like choke-and-pool structure (in one stream) were installed. Two years of post-construction hydraulics data were compared to data collected for two years prior to construction to determine the efficacy of the various stream modifications. Initial evaluations indicated unsuccessful performance of gabion step pools, so after the first year, they were retrofitted with boulders to increase flow depth, restrict discharge, improve flow duration, and create unimpeded connections rather than sudden drops. Variation of lake levels and duration, variability, and depth of stream flow indicated that outlet geometry and lake catchment area should be important considerations when enhancing connectivity for fish migration in ephemeral systems. A narrow, rectangular cross-section was deemed effective for increasing flow depth while decreasing discharge, resulting in increased duration of flows. Catchment area was an effective indicator of a headwater lake’s potential response to connectivity enhancements. Smaller catchments may provide inadequate runoff to sustain minimum storage requirements for enhanced connectivity. Our findings should advance the knowledge of headwater system hydraulics in the Barrenlands and assist in designing future fish habitat compensation projects on similar systems.

Effects of Step Pool Porosity upon Flow Aeration and Energy Dissipation on Pooled Stepped Spillways

AbstractThe hydraulics of stepped spillways with flat steps has been studied for the last three decades, including for embankment dam slopes, but studies of alternative stepped designs are limited. In this study, a pooled stepped spillway was investigated in a relatively large-size facility, and three different pool wall porosities were tested. The flow patterns, the macro- and microscopic air-water flow properties, and the energy dissipation performances were recorded; the results were compared with the flat stepped spillway design for the same chute slope (θ=26.6°). The investigations highlighted a close agreement between air-water flow properties on the configurations in terms of void fraction, turbulence levels, bubble count rate, and chord sizes. The interfacial velocity distributions showed larger interfacial velocity on the pooled step configurations of approximately 5–10% linked with a reduced flow depth. On the porous pooled stepped spillways, the interfacial velocities within the cavity highligh...

029 Rooster sperm plasma membrane protein and phospholipid organization and reorganization attributed to cooling and cryopreservation

Cholesterol to phospholipid ratio is used as a representation for membrane fluidity, and predictor of cryopreservation success but results are not consistent across species and ignore the impact of membrane proteins. Therefore, this research explored the modulation of membrane fluidity and protein organization of rooster sperm during cooling and cryopreservation using the fluorescent stains Merocyanine 540 (phospholipid order) and Rhodamine 640 (protein order). Rooster semen samples (3 roosters and breeds per pool; 4 pools total) were diluted in Lakes Low Temperature (LLT) and Sasaki diluent and cooled to 5 °C over 30 min (Step 1). The samples were then diluted with the penetrating cryoprotectants (CPA) glycerol (LLT) or methylacetamide (Sasaki), equilibrated for 30 min (Step 2), loaded into straws and cryopreserved. The cryopreservation and thawing (Step 3) was performed using the optimized procedures for each diluent. Computer automated semen analysis (CASA) was used to analyze motility characteristics and flow cytometry was used to analyze plasma membrane phospholipid, protein organization and plasma membrane integrity (PMI) at each step. Percent values from CASA and flow cytometry fluorescent medians (FM) were transformed using arcsine and Box–Cox, respectively. The GLM ANOVA was used to detect differences in motility characteristics and membrane organization and included the effects of cryopreservation step and rooster semen pool. In the total motility analyses cryopreservation diluent/CPA was a significant ( P P P P P Source of funding: USDA-ARS-NCGRP National Animal Germplasm Program. Conflict of interest: None declared. phil.purdy@ars.usda.gov

Comparison of different methods to predict the mean flow velocity in step-pool channels

Steep mountain streams have irregular bed topography, where the mean flow velocity is heavily affected by the coarsest bed components and by their arrangement to form step pools, cascades, and rapids. According to literature findings the mean flow velocity is often related with water discharge, channel slope, and grain-size related variables through power relationships. Several approaches consider dimensionless hydraulic geometry terms to develop the analysis over a wide range of channel sizes and hydraulic conditions. The aim of this research is to test the performance of some literature formulas to directly compute the mean flow velocity (V) in step-pool sequences. The study area deals with two fish ladders located in the Vanoi torrent (Trento Province, Italy), which were built by mimicking the steppool morphology. Three reaches were selected to cover different channel slopes (2.6-10%). Data collection entailed three main phases: (1) topographical surveys, (2) granulometric analysis, and (3) flow discharge measurements (salt dilution method). Geometric and hydraulic variables were measured for the following step-pool cross sections: step head, pool center, and intermediate position between pool end next step. Particular attention has been reserved to determine the effective mean flow velocity over the whole path of each step pool sequence. The performance of different literature equations to predict V has been verified. The relations have been shared in three groups: dimensional (V), dimensionless with respect to the grain size (V*) or to a combination of grain size and slope (V**). In general, the V group of equations has produced the highest errors between computed and measured values. The dimensionless V*, V** groups have shown the best performance. In particular the V* equations, which use unit discharge and channel slope, have provided the better fitting, and the lowest root mean square error. The results highlight the difficult to estimate flow velocity in step-pool sequences, and the attitude of this channel-bed morphology to be highly dissipative. The good performance of some dimensionless equations to predict V could also support the hydraulic designer in case the ‘morphological rebuilding’ of mountain creeks is opportune. Further analyses are required to better understand the flow behavior in streams where very rough bed forms and hydraulic drops are the primary sources of flow energy dissipation.

Comparison of different methods to predict the mean flow velocity in step-pool channels

Steep mountain streams have irregular bed topography, where the mean flow velocity is heavily affected by the coarsest bed components and by their arrangement to form step pools, cascades, and rapids. According to literature findings the mean flow velocity is often related with water discharge, channel slope, and grain-size related variables through power relationships. Several approaches consider dimensionless hydraulic geometry terms to develop the analysis over a wide range of channel sizes and hydraulic conditions. The aim of this research is to test the performance of some literature formulas to directly compute the mean flow velocity (V) in step-pool sequences. The study area deals with two fish ladders located in the Vanoi torrent (Trento Province, Italy), which were built by mimicking the steppool morphology. Three reaches were selected to cover different channel slopes (2.6-10%). Data collection entailed three main phases: (1) topographical surveys, (2) granulometric analysis, and (3) flow discharge measurements (salt dilution method). Geometric and hydraulic variables were measured for the following step-pool cross sections: step head, pool center, and intermediate position between pool end next step. Particular attention has been reserved to determine the effective mean flow velocity over the whole path of each step pool sequence. The performance of different literature equations to predict V has been verified. The relations have been shared in three groups: dimensional (V), dimensionless with respect to the grain size (V*) or to a combination of grain size and slope (V**). In general, the V group of equations has produced the highest errors between computed and measured values. The dimensionless V*, V** groups have shown the best performance. In particular the V* equations, which use unit discharge and channel slope, have provided the better fitting, and the lowest root mean square error. The results highlight the difficult to estimate flow velocity in step-pool sequences, and the attitude of this channel-bed morphology to be highly dissipative. The good performance of some dimensionless equations to predict V could also support the hydraulic designer in case the ‘morphological rebuilding’ of mountain creeks is opportune. Further analyses are required to better understand the flow behavior in streams where very rough bed forms and hydraulic drops are the primary sources of flow energy dissipation.

Investigating step‐pool sequence stability

Step‐pool units, common features in steep, narrow streams, are highly dynamic systems that adjust during high flows and active sediment transport conditions. Consecutive step pools in a reach form a step‐pool sequence, and though these features are prevalent in nature, quantifying the stability of such systems is challenging. This study focuses on the statistical relationships between 445 stable sequences of three or more steps and nine geometric and resistance‐based parameters. Step sequence stability is parameterized through a sequence stability parameter, a metric of how long a given step‐pool sequence exists relative to the total run time. Despite variability in the strength of the statistical relationships, the results allow identification of dominant trends between sequence stability and both roughness and geometric parameters. The sediment concentration ratio provides a means for comparing the relative strength of these statistical relationships. As sediment concentration ratios increase, the relations between flow resistance parameters and step stability change from dominantly inverse to dominantly direct. Thus, at higher sediment concentration ratios, sequence stability is more likely to increase with flow resistance, but the reverse is more likely at low sediment concentration ratios. A connection exists between bed morphology, flow resistance, and stability. For the majority of stable step sequences, resistance parameters increase over time, indicating some sequences adjust their geometry to increase stability during flood events. The influence of sediment concentration over the stability‐geometric and stability‐resistance parameter relationships may enable the use of the sediment concentration ratio as a predictor of step sequence stability during a flow event.

Volunteer Post-Construction Monitoring of Stormwater Attenuation in a Regenerative Step Pool Storm Conveyance System Retrofit

Volunteer Post-Construction Monitoring of Stormwater Attenuation in a Regenerative Step Pool Storm Conveyance System Retrofit

Geomorphic response of a headwater channel to augmented flow

Since 1974, flow releases from Long Draw Reservoir have increased annual peak flows on La Poudre Pass Creek, Colorado, from ~ 5.6 m 3/s to > 8.4 m 3/s. The creek drains 61 km 2 and channel morphology varies from step–pool to pool–riffle. Comparison of five channel reaches along the creek to channel reaches along neighboring rivers without flow regulation indicates that channel width has increased by as much as a factor of three along La Poudre Pass Creek. Width-to-depth ratio has also increased, the bed material in step–pool channel reaches has coarsened, and residual pool volumes have increased in pool–riffle channel reaches. Pool–riffle channel reaches have undergone the greatest change in response to flow augmentation. Although discharge has increased consistently for all five channel reaches, morphologic response varies in relation to gradient and channel morphology, making it impractical to precisely predict a priori the magnitude of channel response to flow augmentation.

Do channel restoration structures promote ammonium uptake and improve macroinvertebrate-based water quality classification in urban streams?

Comparisons of ammonium uptake parameters in restored and unrestored urban streams suggest that sufficient light penetration to areas where hard substrates have been installed should be an important management consideration to enhance biofilm accumulation and subsequent ammonium removal to the streambed. We studied ammonium uptake parameters and macroinvertebrate communities in 3 types of restoration structures (riffle, cross vane, and step pool) in restored streams and in unrestored urban streams in Greensboro, North Carolina, USA, where urbanization has led to high nutrient concentrations, degraded channel conditions, and low biotic diversity in streams. Restored streams had a significantly higher percentage of large substrates (boulder, cobble, and gravel) and less canopy cover compared to unrestored streams (P = 0.029; t-test), providing substrates and sufficient light penetration for biofilm growth. Benthic chlorophyll a was higher in restored compared to unrestored streams. Significantly shorter ammonium uptake length (P = 0.02) was observed in restored compared to unrestored sites. This effect was probably related to greater biofilm development and therefore more assimilation sites for removal of ammonium from the stream water. Differences in uptake velocity (P < 0.07) and areal uptake (P < 0.06) were not significant. Despite the shorter ammonium uptake length in restored streams, there was little improvement in measures of macroinvertebrate-based water quality classifications between restored and unrestored streams (P = 0.545). Because this study was completed 2 years postrestoration, continued, longer-term monitoring of restored streams is needed for full evaluation of the effects of the restoration approaches used in these streams.

Reading the bed morphology of a mountain stream: a geomorphometric study on high-resolution topographic data

Abstract. High-resolution topographic data expand the potential of quantitative analysis of the earth surface, improving the interpretation of geomorphic processes. In particular, the morphologies of the channel beds of mountain streams, which are characterised by strong spatial variability, can be analysed much more effectively with this type of data. In this study, we analysed the aerial LiDAR topographic data of a headwater stream, the Rio Cordon (watershed area: 5 km2), located in the Dolomites (north-eastern Italy). The morphology of the channel bed of Rio Cordon is characterised by alternating step pools, cascades, and rapids with steps. We analysed the streambed morphology by means of ad hoc developed morphometric indices, capable of highlighting morphological features at a high level of spatial resolution. To perform the analysis and the data interpolation, we carried out a channel-oriented coordinate transformation. In the new coordinate system, the calculation of morphometric indices in directions along and transverse to the flow direction is straightforward. Three geomorphometric indices were developed and applied as follows: a slope index computed on the whole width of the channel bed, directional variograms computed along the flow direction and perpendicular to it, and local anomalies, calculated as the difference between directional variograms at different spatial scales. Directional variograms in the flow direction and local anomalies have proven to be effective at recognising morphologic units, such as steps, pools and clusters of large boulders. At the spatial scale of channel reaches, these indices have demonstrated a satisfactory capability to outline patterns associated with boulder cascades and rapids with steps, whereas they did not clearly differentiate between morphologies with less marked morphological differences, such as step pools and cascades.

Ecological and Hydraulic Studies of Step-Pool Systems

The ecological and hydraulic features of step-pool systems are studied by field investigation, measurement, sampling, and analysis. The study is done on Shengou and Jiuzhai Creeks, where step-pool systems have developed, Fork Gully, where a step-pool system is developing, and the Jinsha River and Jiangjia and Xiaobaini Ravines, where there is no step-pool system. Boulders, cobbles, and gravel tightly interlock and form the steps having an inherent stability that only extreme floods are likely to disturb. Gravel and sand deposit in the pools behind the steps. These steps and pools provide high diversity of habitat for the stream biocommunity. The density of benthic macroinvertebrates in streams with step-pool systems is several 100 times higher than neighboring streams without step-pool systems. A new habitat diversity index is proposed considering the spatial distribution of various substrates, velocity, and water depth. The study reveals that the biodiversity of benthic macroinvertebrates increases with habitat diversity. Measurements with a specially designed instrument were done to study the development of step-pool systems and its effects on resistance to the flow and stream bed stability. A step-pool system maximizes the flow resistance and protects the bed sediment from erosion. Thus, the riverbed and bank slope are stabilized. The development degree of step pools is proportional to the streambed slope. The bed resistance increases with the development degree of step pools. Riverbed inertia represents the stability of the streambed. The development of step-pool systems greatly increases the riverbed inertia, and, therefore, maximizes streambed stability.

Spatio‐temporal variations of hyporheic flow in a riffle‐step‐pool sequence

AbstractSubsurface flow in streambeds can vary at different scales in time and space. Recognizing this variability is critical for understanding biogeochemical and ecological processes associated with the hyporheic zone. The aim of this study was to examine the variability of hydraulic conductivity (K), vertical hydraulic gradients (VHGs), and subsurface fluxes, over a riffle–step–pool sequence and at a high spatio‐temporal resolution. A 20 m reach was equipped with a network of piezometers in order to determine the distribution of VHGs and K. During a summer month, temporal variations of VHGs were regularly surveyed and, for a subset of piezometers, the water level was automatically recorded at 15 min intervals by logging pressure transducers. Additionally, point‐dilution tests were carried out on the same subset of piezometers. Whereas the distribution of vertical fluxes can be derived from K and VHG values, point‐dilution tests allow for the estimation of horizontal fluxes where no VHG is detectable. Results indicate that, spatially, VHGs switched from upwelling to downwelling across lateral as well as longitudinal sections of the channel. Vertical fluxes appeared spatially more homogeneous than VHGs, suggesting that the latter can be a poor indicator of the intensity of flow. Finally, during flow events, some VHGs showed little or no fluctuations; this was interpreted as the result of a pressure wave propagating from upstream through highly diffusive alluvial sediments. Copyright © 2009 John Wiley &amp; Sons, Ltd.

LiDAR-derived slopes for headwater channel network analysis

Abstract Slope is one of the most important distinguishing features for channel morphology. Variations in the computation of slope from a digital elevation model can affect a wide range of hydrogeomorphically derived applications. We compare different methods for computing channel slope using LiDAR-derived digital terrain models (DTMs) with varying resolutions. We chose a headwater basin of the Eastern Italian Alps, characterized by a dense ephemeral colluvial network and a main alluvial channel as our study area. The identified alluvial morphologies are characteristic of steep mountain streams, namely, cascades and step pools. Field surveys were carried out along the main channel and in some small tributaries. Results indicate that a single method for slope calculation cannot estimate channel slope at the hydrographic network scale. The differential geometry approach for slope calculation tends to overestimate field-surveyed channel slope values for all the DTM resolutions (1, 2, 5 m). When a trigonometric approach for slope calculation is applied, 2 and 5 m DTM resolutions give more consistent results. Nevertheless, a reliable channel slope can be derived from a DTM with an appropriate resolution by choosing a suitable method only after considering the channel width.

Bed morphology and generation of step–pool channels

AbstractFlume experiments have been carried out to study the formation processes and the bed morphology of step–pool channels. From the experiments different step types and step configurations could be distinguished depending on the stream power. These step types can be seen as an image of the generation mechanisms of step–pool systems. These results suggest that the bed roughness geometry develops towards a condition that provides the maximum possible bed stability for a given grain size distribution. In contrast to a variety of other studies, antidunes did not contribute to the generation of the step structures. However, the data of the presented study fits well into the region of antidune formation proposed by Kennedy for sand‐bed rivers. This observation points out that step–pool field‐data located in the Kennedy region do not inevitably prove that antidunes played a role in step development. It is rather proposed that in Kennedy's region of antidune formation there exist hydraulic conditions where the flow resistance is maximized. It is suggested that such maximum flow resistance is associated with an optimal distance between the bedforms and their height, independently of whether these are antidunes in sand‐ and gravel‐bed rivers or step–pool units in boulder‐bed streams. The considerations of the Kennedy region of antidune formation and the analysis of planform step types depending on stream power both suggest that steep channels have a potential for self‐stabilization by modifying the step–pool structure towards a geometry that provides maximum flow resistance and maximum bed stability. Copyright © 2008 John Wiley &amp; Sons, Ltd.

Identification of steps and pools from stream longitudinal profile data

Field research on step–pool channels has largely focused on the dimensions and sequence of steps and pools and how these features vary with slope, grain sizes and other governing variables. Measurements by different investigators are frequently compared, yet no means to identify steps and pools objectively have been used. Automated surveying instruments record the morphology of streams in unprecedented detail making it possible to objectively identify steps and pools, provided an appropriate classification procedure can be developed. To achieve objective identification of steps and pools from long profile survey data, we applied a series of scale-free geometric rules that include minimum step length (2.25% of bankfull width ( W b)), minimum pool length (10% of W b), minimum residual depth (0.23% of W b), minimum drop height (3.3% of W b), and minimum step slope (10° greater than the mean slope). The rules perform as well as the mean response of 11 step–pool researchers who were asked to classify four long profiles, and the results correspond well with the channel morphologies identified during the field surveys from which the long profiles were generated. The method outperforms four other techniques that have been proposed. Sensitivity analysis shows that the method is most sensitive to the choice of minimum pool length and minimum drop height. Detailed bed scans of a step–pool channel created in a flume reveal that a single long profile with a fixed sampling interval poorly characterizes the steps and pools; five or more long profiles spread across the channel are required if a fixed sampling interval is used and the data suggest that survey points should be located more frequently than the diameter of the step-forming material. A single long profile collected by a surveyor who chooses breaks in slope and representative survey points was found to adequately characterize the mean bed profile.

The role of morphology on the displacement of particles in a step–pool river system

Although studies of sediment transport in steep and coarse-grained channels have been more numerous in recent years, the dynamics of sediment transport in step–pool river systems remain poorly understood. This paper investigates displacements of individual clasts through Spruce Creek (Québec, Canada), a classic step–pool channel, and the effects of the channel morphology on the path length of the clasts. Passive integrated transponder tags (PIT) were used to track the displacement of 196 individual particles over a range of discharges including the bankfull stage. Clasts were tracked after five sequences of flood events. The results showed that the distance distributions match a two-parameter Gamma model. Equal mobility transport occurs for the particle size investigated during each sequence of flood events. Mean travel distance of the clasts can be estimated from excess stream power, and the mobility of the clasts is more than an order of magnitude less than the model reported in riffle–pool channels. The dominant morphological length scale of the bed also controls the path length of the clasts. These results confirm some preliminary observations on sediment transport in step–pool channels.

Fish fauna of Chalakudy River, part of Western Ghats biodiversity hotspot, Kerala, India: patterns of distribution, threats and conservation needs

Seventy one fish species belonging to 27 families and 50 genera were collected from various sampling sites spread along the upstream and downstream parts of the Chalakudy river, in the Western Ghats biodiversity hotspot in Kerala, South India. Cyprinids were the most dominant group represented by 24 species belonging to 13 genera, followed by the catfishes of the family Bagridae (7 species from 2 genera) and loaches belonging to the family Balitoridae (5 species from 5 genera). Of the 71 species, 4 belong to the ‘critically endangered’ (CR) and 16 to the ‘endangered’ (EN) category. 3 of the critically endangered species are strictly endemic to the Chalakudy river system. We also collected 5 exotic species several of them categorized as potential pest. Fish species richness (FSR) showed a direct relationship with increasing stream order with maximum values observed in the downstream/low elevation regions. Considerable differences were observed in the FSR between different microhabitats with maximum values observed in bed rock and pool riffle and the lowest in regimes followed by step pools. Diversity and species richness decreased with increasing altitude and higher diversity and species richness were observed in sampling sites located between 0–200 m above sea level. A definite temporal partitioning was also observed with regard to FSR with higher richness during the day time and the lowest observed at dawn. Indiscriminate collection of endemic and threatened ornamental fish species for export, exploitation of endangered food fishes by forest-dwelling communities and local fisher folk, pollution, loss of riparian cover, damming and exotic species threaten the rich ichthyofaunal diversity of Chalakudy River. An urgent need exists for studying the life history traits and demography of the most important endemic and threatened fishes, as lack of information on these aspects have significantly affected conservation efforts. The implementation of strong management strategies like fixing total allowable catch (TAC), and the use of quotas are needed to reduce the overall collection pressure with regard to the trade in native ornamental species from the river.

Dissipative analogies between a schematic macro‐roughness arrangement and step–pool morphology

AbstractWe investigate flow resistance developed by macro‐roughness represented by pebbles positioned on a granular layer according to a regularly spaced stripe pattern on steep bed slopes. Flume experiments under various geometrical and hydraulic conditions are carried out and interpreted by means of a theoretical model. Results show that flow resistance reaches a maximum and is due mainly to form drag when the spacing between macro‐roughness stripes is about 10 times the average macro‐roughness height. A statistical analysis based on various field observations of step–pool geometry underlines that this spacing appears to be one of the most frequent occurring in step–pool bed morphology sequences. Comparison between the present results and flow resistance evaluated for step–pools reproduced in the laboratory and observed in the field suggests that step–pool streams are characterized by a bed geometry able to develop the maximum flow resistance. Finally, a criterion is obtained to estimate flow resistance developed by natural step–pool streams when a formative flow discharge occurs based on geometric quantities only. Copyright © 2007 John Wiley &amp; Sons, Ltd.

Step–pool formation models and associated step spacing

AbstractStep–pool bedforms develop under conditions of high flow rates and high sediment influx and transport rates. Step setting events are capable of mobilizing and rearranging the large step‐forming grain sizes. Thus, observations of the processes of step formation are limited to flume experiments. The results of laboratory experiments involving active transport of widely sorted sediment, including large grains whose deposition readily formed steps, are presented. Over 350 step formations were observed and documented, creating a large data set for detailing and evaluating step formation processes.Existing step‐formation models focus on step development from antidunes, particle clusters, transverse ribs or depositional berms, or from a hydraulic regime that includes cascade flow and hydraulic jumps. These models were developed using field measurements of artificial grade control structures or stable step–pool systems during low flow conditions, or degradational flume experiments where flows elucidated the step–pool bedform in the channel profile but did not create equilibrium transport of the large, step‐forming clasts.This research evaluates the pre‐existing formation models and proposes three new models: the rough bed, exhumation and dune models. The rough bed model was dominant throughout the experiments, followed by the exhumation model and then the dune model. A single model was not responsible for all step formations. The rough bed, exhumation and dune models all influenced step development, and every step sequence was created through a mixture of formation processes. This research generated a large data set, where each step spacing is paired with the corresponding step formation model, making it possible to identify a distinct exclusion zone length for each model. Thus, given knowledge of the step formation model, the minimum distance to the next step downstream can be predicted. From the final spacing of a step sequence, it may be possible to deduce the likely formation mechanisms active during a step resetting event. Copyright © 2007 John Wiley &amp; Sons, Ltd.

Growing Green Roofs, City by City

Green roofs—rooftops that are partially or completely covered with vegetation growing in soil medium over a waterproof membrane—have gained momentum over the past six years as building owners recognize their advantages over conventional roofing in terms of better energy efficiency and reduced rain runoff. Now local governments are exploring incentives for moving the practice into the mainstream. A look at cities that are leading the country in green roof coverage reveals a growing range of policy tools.