Open Collections

UBC Undergraduate Research

Monitoring and Mitigating Reed Canary Grass in Stanley Park, BC Bates, Hannah; Bozik, Marija; Pan, Chunyu; Thormeyer, Markus 2021-04

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


52966-Bates_H_et_al_ENVR_400_Monitoring_reed_2021.pdf [ 1.65MB ]
JSON: 52966-1.0398156.json
JSON-LD: 52966-1.0398156-ld.json
RDF/XML (Pretty): 52966-1.0398156-rdf.xml
RDF/JSON: 52966-1.0398156-rdf.json
Turtle: 52966-1.0398156-turtle.txt
N-Triples: 52966-1.0398156-rdf-ntriples.txt
Original Record: 52966-1.0398156-source.json
Full Text

Full Text

AbstractReed canary grass (Phalaris arundinacea), abbreviated RCG, is an invasive species thatthreatens native flora and fauna via competition and alterations to the abiotic environment. In thepast few years, this invasive plant has rapidly spread throughout parks in Vancouver, BC,including Stanley Park, which is bordered by downtown, English Bay and the Burrard Inlet. TheStanley Park Ecology Society (SPES) strives to uphold the ecological integrity and biodiversityof Stanley Park; thus, being able to monitor and manage the spread of RCG properly is of highinterest and priority. Our team will assist in their mission by providing the following deliverablesin this report: 1) an up-to-date map of RCG abundance and distribution in the park, 2) a mapshowing areas where RCG growth and invasion is most susceptible based on abiotic,environmental factors, and 3) an RCG management plan including recommendations. Thesedeliverables will help SPES locate areas of high priority for active management, understandwhere invasions are most likely to occur, and decide which management and/or monitoringtechniques would work best for each area. The objectives satisfied by our project will gearinvasive management practices in the park towards a customized plan that will increase theeffectiveness of restoration approaches and ultimately improve the parks’ biodiversity.The invaded areas are categorized as low, medium, or high density, which wasdetermined by their current abundance and distribution within Stanley Park from 2019-2021.Distribution data was collected by TREK students, SPES volunteers, and members of our team.Reed canary grass was found to persist in high densities near fresh-water bodies and trails and inlower densities further away from water bodies. This shows that the presence of fresh water has alarge impact on the density of RCG. Susceptible areas were determined by predicting futurebehaviours of the species, based on abiotic factors such as sunlight availability, proximity topotential carriers, and hydrology. The data used to model the susceptibility included distancefrom water bodies, distance from trails, and LiDAR canopy cover. Highly susceptible areas werefound in similar locations as high density areas, however not always near trails, possibly as areasaway from trails were not fully surveyed in the field. RCG is most likely to continue spreadingalong the shorelines of Beaver Lake and Lost Lagoon (two large watersheds within Stanley Park-illustrated in light blue in Figure 2 under ‘Results’), as well as to forested areas with large breaksin the canopy. The growth patterns of RCG based on our research indicate that water and sunavailability are most influential when considering the spread of the invasive species.2We identified 22 reports and management plans published by global institutes andexisting literature. Then we analyzed the materials by conducting a simple qualitative analysis.These methods provided us with the baseline to create specific management techniques fordifferent sizes and RCG densities of patches which were outlined (Figure 4), with mowingrecommended for larger areas and a combination of hand-pulling, mulching, and shadingrecommended for smaller areas. Aquatically-approved herbicide would also be a very effectivecontrol method, however this needs to be approved by the Vancouver Park Board. Theseobjectives aim to decrease the spread of Reed canary grass within Stanley Park and increase thepark’s biodiversity, following the consideration and implementation of the management plan bythe Stanley Park Ecology Society.3Table of ContentsAbstract 2Introduction 5Stanley Park 5Reed Canary Grass 6Characteristics 6Ecology 8Environmental factors affecting RCG spread 9Study Objectives 9Methods 10Map 1 11Map 2 12Management Plan 13Results 15Recommendations 20Available Management Strategies 20Mechanical 20Cultural 22Chemical 23Biological 24Management Plans for Stanley Park 24Non-density zone (0) 25Small area with low-density (1, S) 26Small area with medium-density (2, S) 26Small area with high-density (3, S) 26Large area with low-density (1, L) 27Large area with high-density (3, L) 27Additional considerations 28Conclusion 29About the Authors 30References 31Appendix 354IntroductionThis report will provide the Stanley Park Ecological Society (SPES) with baselines onReed Canary grass data, monitoring plans, and management recommendations for mitigating thespread of the invasive reed canary grass, Phalaris arundinacea (RCG), in Stanley Park. As oneof the world’s largest natural urban parks, Stanley Park provides the city of Vancouver with aneasily accessible greenspace. SPES’s role is to uphold the ecological integrity and biodiversity ofthe park by understanding the impacts of the environmental stressors present and managing themadeptly while also prioritizing the needs of park patrons. Recent invasions of RCG in StanleyPark greatly threaten the health of the terrestrial and freshwater ecosystems, resulting in a needfor immediate action by utilizing the recommendations in this report.Stanley ParkStanley Park is bordered by downtown Vancouver and the Salish Sea, and is situatedwithin the unceded territory of the xʷməθkwəy̓əm (Musqueam), Skwxwú7mesh (Squamish)Nation, and Səl̓ílwətaʔ/Selilwitulh (Tsleil-Waututh) Nation. Three main ecosystems interactwithin the park; the terrestrial ecosystem, the freshwater ecosystem, and the intertidal ecosystem,housing an estimated 1500 native species across all the domains of life inhabiting the park(SPES, 2020). This biodiversity, along with many outdoor activities and programs available inthe park, makes it an extremely popular area for the citizens of Vancouver.As a recreational urban park, Stanley Park is constantly experiencing disturbances in theform of pollutants, climate change, recreational activities by park patrons, invasive species, androutine vegetation management, threatening the ecological integrity and biodiversity of the park.SPES has a duty to protect the park’s ecological integrity and biodiversity by raising publicawareness of natural ecosystems through education, research, and conservation, playing astewardship role in Stanley Park.SPES works closely with the Vancouver Park Board to manage and preserve the park’snatural features. SPES also oversees the Special Invasives Removal Team (SIRT), whichmanages the 97 invasive species present in the park (SPES, 2020). Currently, SPES and SIRT arefocusing on removing RCG, and have created several management strategies such as manuallydigging up the whole plant, mowing at least five times a year, covering the infestation areas, and5applying herbicide (SPES, 2013). These pre-existing strategies and locations are integrated andoutlined both within our maps (hatchback areas in Figure A1 & A6) and within the managementreport by expanding on practices already in place. For a complete understanding of how ourmanagement actions will mitigate the spread of RCG, we provide information on RCGcharacteristics, ecology, and the environmental factors which influence its spread in thefollowing section.Reed Canary GrassCharacteristicsNative to Europe and Asia, RCG is a perennial bunchgrass, which is classified as an‘invasive’ in North America (DiTomaso & Kyser, 2013) and ‘exotic’ in BC (Gov. BC, 2021).RCG grows in wet to damp soils, thriving in wetlands, riparian areas, marshes, and peatlands(DiTomaso & Kyser, 2013; Waggy, 2010). This grass is frost tolerant and drought tolerant;however, it is hypersaline intolerant and shade-intolerant (Waggy, 2010).The aboveground body of RCG is about 1-2 m tall, composed of hairless 1-2 cm diameterhollow stems (Figure 1). Flat, hairless, bright green leaf blades protrude at 45-degree angles fromthe stem. These grasses form stands made up of large bunches which can be up to 1 m in width.The grass forms a dense rooting system belowground, and new rhizomes mostly originate belowthe soil surface from other rhizomes, or from buds in the axils of the leaves on abovegroundshoots. Rhizomes growing in soils with higher water contents will be larger (Waggy, 2010). RCGis dormant in the winter, losing colour and becoming persistent straw-coloured stalks (Figure 2).The aboveground body begins to grow again in early spring, with peak growth occurring inmid-June. Peak rhizome production occurs in August and is associated with decliningaboveground growth (Waggy, 2010).6Figure 1. Mature reed canary grass (long green leaf blades), photographed in October 2020 nearLost Lagoon, Stanley Park, by Chunyu Pan.Figure 2. Dormant reed canary grass (light-coloured stalks), photographed in early March 2021near Lost Lagoon, Stanley Park, by Hannah Bates.Inflorescences originate from the tip of the stem and turn from green to purple in fullbloom with flowering periods spanning from June to July (Waggy, 2010). RCG iswind-pollinated and produces a large quantity of pollen during the summer months (Waggy,2010). After seed production, inflorescences turn straw-coloured and spread slightly. Fruitingoccurs in late July or early August, with a single fruiting body producing up to 600 seeds7(Waggy, 2010). After the seeds are dispersed, the reproductive body dies back to the upperleaves.RCG spreads via seed dispersal or vegetative reproduction, where detached stems orrhizomes can establish themselves as a new grass bunch through direct contact with bare soil(Waggy, 2010). All parts of the plant float, facilitating dispersal via vegetative reproduction infreshwater bodies. Seeds can be dispersed through wind, gravity, water, or sticking to the exteriorof animals, people, and tools/equipment (King County Gov, 2015). Seedlings can emerge just8-10 days after seed planting but can stay viable in the seed bank for up to 4 years. Seeds aremore likely to germinate with increased light and soil saturation (Waggy, 2010).EcologyWhen managing RCG, imminent interventions are necessary because unmanaged RCGcan outcompete native species and create unwanted monocultures by forming thick impenetrablemats (Waggy, 2010). These mats can also physically change ecosystem hydrology by chokingstreams and changing soil hydrological and chemical properties through plant-soil feedbackprocesses (Jacinthe et al., 2010; Perkins & Nowak, 2012). This environmental manipulation cancreate a positive feedback loop where the conditions created to result in the displacement ofnative species, creating more room for RCG to invade, establish itself, and form a thickmonoculture mat of grass (Schooler et al., 2006).RCG also creates monocultures by allocating most of its biomass to the abovegroundbody in the first two years of growing: RCG quickly secures light and shades out native species(Adams & Galatowitsch, 2006). This reduced native plant diversity decreases low canopycomplexity and arthropod biodiversity (Weilhoefer et al., 2016). Arthropods are essential interrestrial ecosystems, acting as pollinators and a large component of food webs, thereforedecreased diversity negatively affects ecosystem health and stability through a myriad ofcascading effects (Hallmann et al., 2017). RCG invaded areas appear to have no immediateeffects on the total abundance or richness of small mammals or birds, but evidence of shifts incommunity composition has been observed (Sypreas et al., 2010).8Environmental factors affecting RCG spreadThere are several biotic and abiotic/environmental factors that can facilitate the spread ofRCG. The broad-ranging dispersal capabilities and the long period of seed viability, togetherwith the quick seedling responses to optimal emergence conditions, make RCG excellent atcolonizing frequently disturbed sites (Kercher & Zedler, 2004). Pollution, fragmentation causedby roads and trails, regular vegetation management, and recreational activities by park patrons allact as disturbances to the park which can allow RCG to establish itself. Proximity to waterbodies also facilitates the spread of RCG. RCG thrives in more saturated soils, and access tofreshwater bodies greatly increases dispersal ability to all areas in close proximity to the water.Sunlight levels are big determinants in RCG spread as RCG is not shading tolerant and seedgermination rate is increasingly higher as the light intensity grows (Lindig-Cisneros & Zedler2001). Thus, areas with lower canopy densities are more likely to see RCG invasion. A highlevel of nutrients in the field also attracts and facilitates its invasion (Kercher & Zedler, 2004;Lavergne & Molofsky, 2004).Study ObjectivesGiven the invasive nature of RCG,  we have devised the following objectives to reduceand monitor its presence in Stanley Park. The two main objectives of this project are the zonalmaps of the park and a project recommendation for restorative practices in the park. The mapswill serve as a tool to locate and quantify the density of RCG, both presently and projected in thepark’s regions, for both management and educational purposes. The managementrecommendations will be provided to SPES to reduce RCG spread and mitigate ecologicaldamage. These resources can be used to monitor the spread of the grasses, and also theeffectiveness of management techniques in certain areas. This information will be useful to bothSPES and the Vancouver Park Board for self-assessment on how well management techniquesand restorative practices are mitigating the spreading of the grass.● Objective 1: Determine and map which areas within Stanley park are invaded by RCG;ranking them by density and abundance9● Objective 2: Determine and map which areas within Stanley Park are most susceptible toRCG spread● Objective 3: Create a management plan for SPES which ranks areas invaded by RCG bypriority and effectively mitigates the spread of RCGBy meeting all three objectives, we aim to provide SPES with the tools necessary touphold the ecological integrity and biodiversity of the park through the monitoring andmanagement of RCG.MethodsOur methods rely heavily on data manipulation and literature reviews to guide us towardsour final results. GIS (Geographic Information Systems) and supplementary mapping programswere utilized to plot the reed canary grass’ density and presence. The data used was obtainedfrom various sources, listed in further detail below, and compiled into zones where the grassesare most evident. The data is presented in a polygon form rather than point form within ArcGIS(our chosen GIS for the project) to represent these regions rather than represent each individualplant within the park, thus creating clearer boundaries for maintenance. The idea of the zones isto allow easier comparison between 1) the ecosystems found within the high density regions andno/low density regions and 2) the progress of the restorative method within each region as timeprogresses.Additionally, our mapping and analytical techniques are supported by informationcompiled from literature reviews to provide a background understanding of our maps.Knowledge on the frequency of disturbances, ie. mowing, abiotic factors, competition, and theecosystem around the highly invaded areas can offer us insight into the reasoning for the invasiveplant’s spread. We used our own collected data and referencing methodologies from publishedliterature to determine which biotic and abiotic factors were most appropriate to include in ourmodel, as was introduced in ‘Environmental factors affecting RCG spread’ in the introduction ofthis report.The RCG management recommendations were created after consulting literature reviewsand other RCG management plans from areas with similar climates and/or landscapes. These10recommendations take into consideration the resources available (specifically volunteer labour),seasonal and yearly timelines, any restrictions SPES may have on certain management practices,and the relationship between the park’s visitors and the park’s ecology. The methods for each ofour three proposed objectives are underlined below.Map 1To assess the current abundance and distribution of reed canary grass, we created a map(Figure 1) illustrating the affected areas. We performed our data analytics and mapping usingArcGIS Pro version 2.7. The data we used for this map included data collected by students in theTREK program in 2020 and 2021, SPES volunteers in 2019, and by our team in early 2021;therefore, our collective data for Figure 1 indicates the abundance of RCG in Stanley Park from2019 to early 2021.We have separated our methods below by source, as each dataset had slightly differentattribute data and thus had to be processed differently.Table 1. Methods used to create Map 1 according to the data source.InputDataSourceYearCollectedFileFormatMethod DescriptionTREKstudents(collectedusingGPS)2020 -2021Point Data was categorized into ranges of severity ofinvasiveness, from low to high, based on the densityindex and the length of the patch of grass along thetrails. These indexes were calculated by accounting forthe quantitative and qualitative data collected: thelocations with higher density and longer tracks had ahigher index value associated with that point. Thedensity was marked as either ‘sparse’ or ‘dense’ andgiven a weight of 1 or 2, respectively.SPESvolunteers(collectedusing2019 Point The same methods as for the TREK data (above) wasutilized for the SPES 2019 data. Additionally, this data11ArcGISCollectorApp)included points for other non-RCG invasive species,which were manually deleted.HannahandMarija(collectedusingArcGISCollectorApp)2021 Point &PolygonData was divided based on area covered: polygons withareas over 600 m2 remained in polygon form (as theyalready allocate at-risk areas), and areas under 600 m2were configured into points for further analysis andaggregation. The reasoning for this distinction is due tothe variance in data collection between data sets whichrequires us to normalize the data before being able toaggregate it into polygons with the other datasets,following the same methods as those outlined above.Once all point layers were adjusted, they were combined to a single layer. A quantilepercentage (3 classes) was used to create the severity gradient for RCG density (low, medium,high) based on the density indexes, and separated into three separate layers. The quantilepercentage (3 classes) method for classification was used to reduce skew and simplify the visualoutput. For each of these layers, the points were aggregated (within 150 m) to form polygons.These polygons were then buffered to 20 m to account for the areas surrounding the RCGpatches that may have been overlooked or are at risk for RCG spread. The polygons denote therelative severity of RCG density by colour: red for high density, orange for medium density, andyellow for low density (Figure 4).Map 2To locate areas in Stanley Park most susceptible to RCG spread, the following abioticfactors were considered: sunlight availability, undergrowth cover, hydrology, and trail presence.Sunlight availability and undergrowth canopy data layers were extracted from LiDAR data (Cityof Vancouver, 2018), and converted to raster format using ArcGIS Pro. Areas with high canopycover typically have less available sunlight for the undergrowth. Areas with low canopy coverindicate RCG may also be able to thrive there, indicating the absence of roads or buildings.12Hydrology and trail presence were obtained from SPES as shapefiles. These shapefiles werebuffered by 5m, and then converted to raster format. These layers are used to indicate areas nextto water bodies and trails.The central component of map 2 was to create a weighted susceptibility raster layer(Figure A2 & A3), which allocated a weighted index to each of the factors considered whenlooking into the potential distribution and abundance of RCG in the future. We gave the highcanopy factor the highest weight (of 2), as this factor is of high concern to the SPES team, andweighed low canopy, hydrology, and trails at equal weights (of 1).  We used these weightings tocreate a susceptibility raster, which highlights the areas of the park most susceptible to RCGspread based on the above abiotic factors. The ‘Locate Regions’ tool was used to locate areas ofrelative high susceptibility within the raster (Figure A4), and polygons were drawn around areassusceptible to RCG spread based on these regions as well as the underlying susceptibility raster.(Figure 5). The polygons created for Map 1 were included in Map 2 to allow for monitoring ofRCG’s movement in Stanley Park, by comparing its abundance to past years.Management PlanWe identified 22 reports and management plans published by global institutes, such asWisconsin Reed Canary Grass Working Group, and the existing literature as the database. Toanalyze the materials effectively, we conducted a simple qualitative analysis. Qualitative researchnot only enables us to explore essential patterns, phenomena, observations beyond quantitativeresearch (Bartunek & Seo, 2002; Watkin, 2012), it can also help to obtain a more profoundcomprehension to solve our research questions more solidly (Rust et al., 2017). We used NVivo,a Computer-Assisted Qualitative Data Analysis Software (Woolf and Silver, 2018), toqualitatively analyze these files to observe the pros and cons and popularity of RCG managementstrategies. For example, mowing is the most frequently discussed mechanical method, but maynot be the most effective, and can have certain drawbacks.13Figure 3. Screenshot of the node structure for management strategies from NVivo.We immersively read all of the articles and inductively created a node structure (Figure3). We identified five broader control types: chemical, mechanical, biological, cultural, andpreventative, with specific measures listed under each type. All the relevant sentences andparagraphs can be coded to this structure, and the corresponding information will appear afterdouble-clicking any of the icons. Thus, we managed to synthesize and summarize all of the keypoints for each strategy. To further assess the popularity, we input the coding frequencyinformation to Excel to create visualized graphs. Coding frequency refers to the number ofsources discussing a particular strategy: for example, 18 files out of 22 contain a discussionabout chemical control. With the information indicated from the map results, discussions withSPES, and the insights provided by this analysis, we comprehensively providedrecommendations for the management plan of RCG for Stanley Park.14ResultsFigure 4 shows the abundance of RCG in Stanley Park from 2019 to early 2021.High-density areas tend to be near water bodies and walking trails. Low-density areas tend to befound along walking trails, but not as close to water bodies. As shown in Figure A1, there issome overlap between high-density zones and current SPES worksites in most areas, with BeaverLake (Figure A1, top left) having the least overlap with worksites. This shows that futureworksite locations should focus on this area to have the greatest effect on RCG removal, and thisarea may be at higher risk of spreading or intensifying as there are currently few worksites in thearea.It should be noted that some areas were not sampled, and thus could not be included inthis density map, and it is unknown if RCG does persist in these regions. This includes along theSeawall on the east side of the park, and forested areas very far away from walking trails. On asite assessment in March 2021, we found only young and low-density occurrences of RCG alongthe Seawall on the west side of the park. As a result, we are somewhat confident that there arecurrently no high-density areas (potential high priority zones) along other parts of the Seawall,however further surveying is needed to confirm this. As shown in our second map (Figure 5;more detailed in Figure A5, bottom left), certain sections of RCG near the Seawall are likely toexpand in distribution and abundance, as the appropriate abiotic conditions can exist here. As forheavily forested regions away from walking trails, we find it also unlikely that there arehigh-density occurrences, as these areas usually have dense canopy cover (little sunlight) and areaway from spreading mechanisms such as water and humans using the trails. Our susceptibilityraster (Figure A2) also shows these areas as being the least susceptible to RCG invasion/growth.While additional surveying will be needed to confirm this, it is less likely that high-density RCGpatches persist in these areas, and focus should be drawn to those shown in Figure 4.15Figure 4. Map 1, showing reed canary grass density areas in Stanley Park from 2019 to 2021, with red being the highest density. Seeappendix (Figure A1) for detailed sections.16Figure 5 includes areas where reed canary grass is most likely to spread and/or grow (inpurple), based on our susceptibility model. These areas tend to be near water bodies and in areasmore exposed to sunlight (little canopy cover), though not necessarily close to trails. This couldbe due to the fact that while current high-density areas were based on sampling near trails, thesusceptibility model included abiotic factors which might occur away from trails. As shown inFigure A6, some of these areas already exist within current SPES worksites, such as thosearound Lost Lagoon (bottom right). The ‘likely spread’ areas around Beaver Lake (Figure A6,top left), the Seawall (bottom left), and the Prospect Point picnic area (top right) are not withincurrent SPES worksites.As stated previously, there are still some gaps in the surveying range, such as part of theSeawall and forested areas far away from trails. Other, smaller gaps may have also beenunintentionally left out of surveys, such as short sections between surveyed areas. Oursusceptibility raster (Figure A2) indicates where RCG may thrive based on abiotic conditionsand could be used as a tool to help locate future surveying locations for RCG, as well asremotely assess the likelihood of RCG occurring in specific areas, helping fill these gaps.These two figures indicate where management methods for RCG should be appliedwithin Stanley Park, due to current abundance (Figure 4) and predicted spread (Figure 5). Theyshow that RCG has a high presence in the park, and this is anticipated to increase, withoutsuccessful mitigation and management. Figure 6 relates to the assessment of different potentialmanagement strategies for RCG.17Figure 5. Map 2, showing areas (in purple) where reed canary grass is most likely to spread to in Stanley Park. See appendix (Figure A5 & A6) fordetailed sections.18Figure 6. Percentages of control strategies discussed in the literature review (n=22). Differentcolours refer to different broader strategies: Chemical (Red), Mechanical (Blue), Cultural(Green), Biological (Grey), and Prevention (Orange).Figure 6 shows the frequency of the different control methods discussed throughout 22published reports, management plans, and literature. For example, herbicide was discussedwithin 73% of the 22 sources, representing the most popular method. The next two most frequenttechniques are mowing and burning. Cultural and biological control strategies were moderatelydiscussed, while prevention, excavation, and disking were the least frequent. However, thepercentage does not always reflect the effectiveness or actual implementation of certainmanagement strategies. For example, while biological control was discussed by 32% of thesources, the discussion sections stated that there were no effective biological control methods;hence, we cannot conclude biological control type is an effective management technique. Adetailed discussion of each control strategy will be presented in the following section.19RecommendationsThis section will be presented in two parts; the first part is the discussion of the status ofall existing RCG control strategies, and the second part is the specific application of thesestrategies to Stanley Park.1. Available Management StrategiesPreventionPrevention is the most important and cost-effective component of an IntegratedManagement Plan because it prevents the RCG invasion from occurring in the first place. If weare proactive in ensuring the site conditions are resistant to invasion, then the cost would beminimized because fewer control practices would be needed.It is difficult for RCG to invade intact, non-disturbed, healthy wetland ecosystems.Therefore, altering some necessary site conditions to maximize the health and resistance of thehabitat are crucial for prevention. Firstly, establishing a complex canopy to limit sunlightavailability on the soil prevents the germination and establishment of RCG. Even if an area isuninvaded, planting fast-growing trees such as willows at sites with low canopy density isessential to prevent invasion. RCG invasion could be further prevented by enriching the carbonin the soil to reduce nitrogen nutrient level, and then planting a sedge like Carex hystericina,native to Wisconsin and Minnesota, to suppress its growth. Monitoring is also a critical part ofprevention (Tu, 2004). By monitoring the vulnerable sites periodically, such as wetland areas andareas with limited canopy cover, RCG invasion could be identified early and addressed before itspreads to nearby sites or establishes itself as a thick monoculture stand.MechanicalMowingMowing is an effective and popular mechanical strategy. Using mechanical equipment(such as tractors or heavy mowers) to quickly remove a large patch of RCG is suggested becauseRCG is grows in dense stands (Apfelbaum & Sams, 2006; DiTomaso et al., 2013; Kilbride &Paveglio, 1999; Lavergne & Molofsky, 2006). Mowing needs to happen before the maturation20and seed production of RCG. Before or at the stage of anthesis, RCG allocates the majority of itsenergy and resources to the development of seed heads. Therefore, mowing at this time wouldgreatly weaken the vigour of the RCG invasion. It is suggested to mow 5-8 times annually for5-10 years, because the RCG will quickly regrow from its rhizome networks. It is stronglyrecommended to combine mowing with other strategies to maximize the effectiveness. For areaswith extremely high RCG densities, a follow-up herbicide application is heavily documented andrecommended by the literature.BurningBurning is another effective and popular control strategy for large invaded areas, but maybe prohibited in many places due to ecological, environmental, safety, or aesthetic reasons.Prescribed burning is recommended only for wetlands with fire-adapted species, so the nativespecies will grow back after the fire application (Adams & Galatowitsch, 2006; Apfelbaum &Sams, 2016; DiTomaso et al., 2013; Foster & Wetzel, 2005; Reinhardt & Galatowitsch, 2004).Annual burning around late spring or late autumn for 5-6 consecutive years could control andeliminate RCG. Burning in the early spring could have the opposite effect and stimulate RCGinvasion because RCG is an early-growing species and can outcompete native species during thefire disturbance, and their rhizomes may recover in the summer (Lavergne & Molofsky, 2006).Combinations of burning with other control strategies, such as burning prior to herbicideapplications, will give the best results.Hand RemovalHand pulling is only practical and realistic for areas with small infestations because it istime-consuming and labour-intensive (DiTomaso et al., 2013; Lavergne & Molofsky, 2006).Only small stands of RCG can be effectively controlled by hand removal. It is best to pull whenthe grass is immature because it is near impossible to remove every part of RCG once itsrhizomes spread. Mid-spring is suggested as the best time to hand pull because this time of yearis when RCG develops its secondary leaves. The root system needs to be dug with hand tools andthe rhizome must be completely removed to the best of the hand-puller’s ability.21DiskingDisking is an agricultural treatment for soil, usually using heavy tractors to break the hardclods and crusts on the surface. RCG is sensitive to disking treatment (DiTomaso et al., 2013;Kilbride & Paveglio, 1999). However, disking must be combined with other treatments sincedisking alone would not effectively control RCG. A pre-treatment of herbicide application ishighly recommended.ExcavationExcavation is generally not recommended for controlling RCG by the literature.CulturalShadingPlanting fast-growing trees or shrubs can create a shade canopy that would shade out theRCG (Tu, 2004). It can suppress the growth of RCG because sunlight is required for RCG toflourish. It is suggested that planting to establish a multi-layer canopy system would produce thebest results. The following species have so far proven to be effective: alders (Alnus glutinosa),cottonwoods (Populus tremula), willow (Salix alba), and native clover (Trifolium pratense).Deciduous trees may not fully and successfully eliminate the RCG. Another note is that it couldbe difficult to plant new trees where beavers are active; therefore, active monitoring andprotection is required. Since shading can only control the aboveground biomass of RCG and notits rhizomes, it is suggested to use the shading method in conjunction with other strategies.MulchingMulching, also known as covering, is recommended for small infestations, usually 6meters across (Tu, 2004). The covering layers can be fabric or plastic, as long as they are opaqueand prevent photosynthesis. The best time for mulching is the early spring because not manyseed heads are produced. RCG should be removed before placing the covers. There should beseveral layers of anchored covers and the intersections of each mulch should be at least 30 cm.The total area of coverage needs to extend past the area of infestation by at least 60 cm. Covers22should be left for 2 growing seasons and monitoring should continue if there are new infestationsnearby. Mulching can also be combined with other methods, such as shading and mowing.Hydrological ManipulationFlooding can be effective if the water level is raised by at least 80 cm and this level ismaintained for the entire duration of the growing season, or longer (Lavergne & Molofsky,2006). If the flooding is not maintained for a long enough period, RCG has enough tolerance tooutcompete the native species to regrow under the fluctuating hydrological regime. Floodingmay pose a negative effect on native species, and therefore, this method is more appropriate forareas dominated by RCG.GrazingGrazing is potentially a control method but not for wetland settings (DiTomaso et al.,2013). The palatability of RCG is only good when the stems and leaves are young; cattle willgraze on RCG in the late season. Grazing will not eliminate RCG but can be mixed with othercontrol strategies.ChemicalHerbicideUsing herbicide appropriately is proven to control the noxious RCG effectively (Adams& Galatowitsch, 2006; Apfelbaum & Sams, 2016; DiTomaso et al., 2013; Kilbride & Paveglio,1999; Lavergne & Molofsky, 2006; Foster & Wetzel, 2005; Reinhardt & Galatowitsch, 2004).However, it is hard to define and standardize the proper application time. For most cases, fallapplication could be more effective than spring application. Applying herbicide during the fallcan eliminate the grass, while a spring application can result in RCG recovering in the summer.Moreover, an appropriate application rate which depends on different types of herbicide, alsoplays an important role. For example, the aromatic amino acid herbicide has entirely differentusage strategies from the branched-chain amino acid herbicide. Repeated application of herbicidemay be required. Since RCG thrives in wetland environments, herbicide usage can be prohibitedin many cases. Improper use of improper herbicide type can decrease local aquatic ecosystem23functioning by changing pH and introducing pollutants. If the persistence and toxicity of theherbicide are high, the adverse consequence would be worse for the long term health ofecosystems. In some coastal states, like Washington State, only the aquatically-approvedherbicide, such as Glyphosate, is permitted for use with a valid licence. A single herbicide maynot be enough, and many studies have suggested that a combination of herbicide and burning ormowing is double insurance to the success of RCG control. These treatments need to be repeatedfor several years to maximize success rates of RCG removal.BiologicalAll sources stated that there are currently no known biological control options for RCG(DiTomaso et al., 2013; Kim et al., 2006).2. Management Plans for Stanley ParkFour types of RCG zones have been identified: high density, medium density, lowdensity, and non-priority zones. These are separated into two sizes of areas: small (<4000 m2)and large (>4000 m2). Therefore, we provide six specific management plans for each category(Table 2) and these plans are described in detail below. Figure 7 shows an overview of thedifferent management recommendations based on the different site conditions.Table 2. The abbreviations of different infestation combinations.Small Area (S) Large Area (L)Non-Density (0) (0)Low-Density (1) (1, S) (1, L)Medium-Density (2) (2, S) N/AHigh-Density (3) (3, S) (3, L)24Figure 7. Visualization of specific management plans for different invasion conditions.Non-density zone (0)Non-density zones are all the areas not enclosed by a polygon in Map 1 (Figure 4), i.e.,all of the uncolored areas. The management objective is to keep these areas clear of infestations.The principal mechanism is prevention. We recommend checking the canopy conditions of thenon-priority zones in the park once to twice a year and recording the sites with large canopy gapsas potentially vulnerable areas. Initially, these areas may be able to be found remotely via FigureA2, although the canopy data used (from 2018) will become less accurate with time. Also, asStanley Park is a public park, we recommend checking whether there is new infestation whenchecking the canopy cover because the seeds of RCG may be easily carried by park patrons. Werecommend continuing to monitor all of the vulnerable sites once per month, including largecanopy gaps and wetlands. If there are new infestations in the vulnerable sites, assess the prioritylevel and acreage and initiate the corresponding management plans below.25Small area with low-density (1, S)Since it is a small infestation area and the risk of further spread and greater invasion islow, we recommend implementing hand pulling and prevention as a low-cost but effectivecontrol. Hand-pulling is very efficient for small infestation areas and can remove the existingRCG in a short period of time and at a low cost. Hand-pulling could be conducted by groups ofvolunteers with proper instructions, and this type of community volunteering work in StanleyPark can be attractive to nearby high school students. To conduct hand pulling, pull the RCGwith gloves on and dig out the rhizomes (as many as possible), as any roots left in the soil willregrow. We recommend disposing of the materials responsibly by burying all the removed plantparts under 60 cm of soil and cleaning all the tools used. After hand removal, proceed with theprevention measures by monitoring the site conditions once a month and recording if there arenew infestations. If there are large gaps in the canopy, plant trees like willow and alder for futureprevention.Small area with medium-density (2, S)To control small areas with medium density, we will implement hand pulling, mulchingand prevention. Medium-density means that these areas are riskier than low-density for potentialRCG spread in the park, and need more strategies than (1, S) to ensure the effectiveness. Wepropose to introduce mulching to these areas. After safely hand pulling all the RCG, as directedin (1, S), in the small infested areas, we recommend laying enough burlap sack, shade cloth,cardboard or fabric layers, on the treated area. Anchor these layers with overlaps of at least 30cm, and ensure that the total area of mulching surpasses the infestation area by at least 60 cm.Multiple layers overlapping will give the best effects. Then, leave the mulching for two growingseasons and initiate the prevention processes. It is critical to continue monitoring the site andcheck whether any new RCG reaches out of the layers for sunlight.Small area with high-density (3, S)To control small areas with high density, we suggest implementing hand pulling,mulching, shading, prevention, and a contingency plan. High density indicates that the area isfacing a high risk of RCG spread, and with no control, RCG will invade and spread to other areas26in the park quickly. Therefore, apart from hand-pulling and mulching, we propose to implementthe shading method and a special contingency plan if the condition worsens. In this case, wesuggest mixing mulching and shading together for increased effectiveness. Use the willow livestakes to stick through the mulch layers, one stake per burlap sack. This fast-growing willowspecies would typically form shade in 5 years. With both coverings and the shades, the RCG willhave a fairly low chance of regrowing. However, this is a high RCG density area and theconditions are favoured for invasion, leading to unforeseeable situations. We recommendregularly monitoring this area at least once per week, and if the situation becomes out of control,prescribed fire or aquatically approved herbicide may be needed as the contingency plan. Sincethe herbicide and fire uses are not approved by the Park Board, we recommend applying thesemethods for contingency use, especially the aquatically-approved herbicide, as a potential option.Large area with low-density (1, L)To control large areas with low density, we propose implementing mowing, shading andprevention. A large area indicates that the area of infestations is large, and hence, the totalinfestations are too many to conduct by hand removal method. Therefore, we suggestimplementing the mowing technique for quick removal. Use a heavy mower or tractor at leastfive times per year and continue mowing for at least five years. However, Stanley Park will notallow more mowing activities beyond the trail edges. Therefore, we recommend mowing as largeas possible for allowed areas and then hand pulling for the remaining areas. Clean the site asmuch as possible and dispose of the removed plant parts responsibly. We propose to plantfast-growing tree species, such as willow, for shade formation. If the herbicide is permitted foruse, a follow-up herbicide application can help eliminate the RCG. Lastly, continue monitoringthe site afterwards to check whether there are new RCG infestations, as a preventative measure.Large area with high-density (3, L)To control large areas with high density, we recommend implementing mowing, herbicideor burning, and prevention. Given that it is a large area with a high risk of spread, an efficientand effective approach to RCG elimination is needed in these areas. Although Stanley Park doesnot allow mowing other than trail edges, we strongly suggest applying for mowing permissions27in this area since this is the most risky area. We suggest mowing the area first and then applyingprescribed burning or aquatically approved herbicide. Also, if the prescribed fire is impossible tobe permitted, we recommend trying to approve the use of herbicide. If the prescribed fire ispermitted, it needs to happen 5-6 years in a row every autumn. If the herbicide is permitted,follow the instructions of application time and rate provided by the specific brand. In most cases,the combination of mowing and herbicide will need repetitions for at least three growingseasons. After herbicide, we suggest restoring the site by planting the shading trees, willow oralder, and native species. Continue monitoring the site and check whether there are newinfestations.Additional considerationsIn addition to the management plan and priority zones stated earlier, there are specificareas in the park and supplemental management ideas which warrant some consideration. Thefirst area being Lost Lagoon (Figure A5, bottom right), which contains a biofiltration pond thatcatches runoff pollutants from the nearby roads and likely harbours RCG. The biofiltration pondis not physically accessible for performing RCG management techniques and will likely act as apermanent reservoir for Lost Lagoon and nearby areas, making full eradication near impossible.Trails that have been widened should be treated as special areas and are to be closely monitoredfollowing widening. This is due to the increased sunlight availability, which allows dormantRCG seeds in the seed bank to germinate and potentially establish a newly invaded area.Recently restored areas should also be vigilantly monitored for 4 years following initialremoval and management treatments because RCG seeds can stay viable in the seed bankthroughout this duration. Invasive plant restoration project success rates are significantly higherif the project is completed and followed up with the establishment of a native plant community.For this to happen, the native plant community must be monitored and continually managed untilit has established itself as a stable state (Hazelton et al., 2014). Restoration plans should only beinitiated when the resources required for this 4 year duration are secured and available.The mowing regimes could also be altered to mitigate the spread of RCG by reducingcontamination between invaded sites. RCG seeds can stick to mowing equipment and themowers themselves. If possible, we recommend collaborating with the Vancouver Park Board tocreate a plan where all mowing equipment, vehicles, and footwear are fully sanitized between28trails. We also recommend optimizing mowing routes which begin in areas that have little or noRCG and end in areas with higher RCG densities. This reduces the chances of seed transfer fromheavily invaded sites to non-invaded sites. All parts of RCG which are removed from sitesshould also be disposed of through safe and sanitary methods. RCG removed from the groundcan reestablish itself once in contact with bare soil again.Finally, positioning boot brushing stations at the start and end of every trail should beconsidered. These could be paired with infographics explaining their purpose and educating parkpatrons on the prevalence and potential destruction invasive species can cause in ecosystems.Brushing boots would remove sticky RCG seeds from the boots of park patrons, reducing thespread of RCG seeds between trails, while providing information on why it is necessary.Increasing public awareness of the presence and implications of restoration projects is likely toincrease the longevity and success of restoration projects (Crall et al., 2011; Devictor et al.,2010).ConclusionMaps showing the distribution of RCG and its predicted spread, and an assessment ofmanagement strategies and recommendations proposed here, provide effective changes thatSPES can implement to reduce RCG presence within Stanley Park. Map 1 (Figure 4) provides abaseline for current RCG abundance and distribution which can be adapted and added to bySPES, as the abundance of this invasive species will change in the future. This reduces theuncertainty of how to best approach the issue and fills in a pre-existing gap of not havingsufficient RCG abundance data. Map 2 (Figure 5) looks forward to the future of the park’s healthand highlights the areas most at risk for RCG invasion or growth. The current distribution ofRCG, and susceptible areas based on abiotic factors that influence RCG growth, can be used todirect future restoration areas. By exposing these at-risk areas early, preventative measuresand/or monitoring practices can be put in place prior to RCG spreading and causing furtherbiodiversity loss in Stanley Park.29About the AuthorsOur team consists of four final-year Environmental Science majors: Hannah Bates,Marija Bozic, Chunyu Pan, and Markus Thormeyer.Hannah Bates is in the Ecology and Conservation Area of Concentration and has takenmany courses in ecology and plant biology, thus fitting for dealing with an invasive grassspecies. She has also taken two courses in Geographic Information Systems, useful for themapping component of this project.Marija Bozik is also in the Ecology and Conservation Area specialization within theEnvironmental Sciences major. By having experience in fieldwork, GIS, and an educationalbackground in ecology, the aims for this project match up well with the skills and passion she hasrelated to the project.Markus Thormeyer is in the Ecology and Conservation Area of Concentration. He hastaken a Geographical Information Systems course giving him the necessary experience withmapping, and an ecological restoration course where he gained experience with real-worldapplications of restoration ecology, and many conservation courses.Chunyu Pan is in the Land, Air and Water Area of Concentration in EnvironmentalSciences. He has taken relevant biology courses, including plant ecology, life history and plantidentification, to build a fundamental knowledge of the project. He has taken one GIS course thatoffers him a knowledge base for mapping.30ReferencesAdams, C. R., & Galatowitsch, S. M. (2006). Increasing the effectiveness of reed canarygrass (phalaris arundinacea L.) control in wet meadow restorations. RestorationEcology, 14(3), 441-451., S. I., & Sams, C. E. (1987). Ecology and control of reed canary grass (phalarisarundinacea L.). Natural Areas Journal, 7(2), 69-74.Bartunek, J. M., & Seo, M. (2002). Qualitative research can add new meanings toquantitative research. Journal of Organizational Behavior, 23(2), 237-242.doi:10.1002/job.132B.C. Conservation Data Centre (2021). BC Species and Ecosystems Explorer. B.C. Minist.of Environ. Victoria, B.C. Available: (accessed Apr16, 2021).City of Vancouver. (2018). LiDAR 2018. City of Vancouver Open Data Portal., A.W., Newman, G.J., Stohlgren, T.J., Holfelder, K.A., Graham, J., & Waller, D.M.(2011). Assessing citizen science data quality: an invasive species case study.Conservation Letters 4(6): 433-442Devictor, V., Whittaker, R.J., & Beltrame, C. (2010). Beyond scarcity: citizen scienceprogrammes as useful tools for conservation biogeography. Diversity and Distributions16(3): 354-362.DiTomaso, J.M., G.B. Kyser et al. (2013). Weed Control in Natural Areas in the WesternUnited States. Weed Research and Information Center, University of California. 544 pp.Foster, R. D., & Wetzel, P. R. (2005). Invading monotypic stands of phalaris arundinacea: Atest of fire, herbicide, and woody and herbaceous native plant groups. RestorationEcology, 13(2), 318-324., C.A., Sorg, M., Jongejans, E., Siepel, H., Hofland, N., Schwan, H., Stenmans,W., et al. (2017). More than 75 percent decline over 27 years in total flying insectbiomass in protected areas. PLOS ONE, 12(10): e0185809Hazelton, E.L.G., Mozdzer, T.J., Burdick, D.M., Kettenring, K.M., & Whigham, D.F.(2014). Phragmites australis management in the United States: 40 years of methods andoutcomes. AOB Plants 6, plu00131Jacinthe, P., Bills, J. S., & Tedesco, L. P. (2010). Size, activity and catabolic diversity of thesoil microbial biomass in a wetland complex invaded by reed canary grass. Plant andSoil, 329(1-2), 227-238.Kercher, S. M., & Zedler, J. B. (2004). Multiple disturbances accelerate invasion of reedcanary grass (phalaris arundinacea L.) in a mesocosm study. Oecologia, 138(3),455-464. doi:10.1007/s00442-003-1453-7Kilbride, K. M., & Paveglio, F. L. (1999). Integrated pest management to control reedcanarygrass in seasonal wetlands of southwestern Washington. Wildlife Society Bulletin,292-297.Kim, K. D., Ewing, K., & Giblin, D. E. (2006). Controlling Phalaris arundinacea (reedcanarygrass) with live willow stakes: a density-dependent response. ecologicalengineering, 27(3), 219-227.King County Gov. (2015). King County Noxious Weed Control Program Best ManagementPractices Reed Canarygrass. King County Gov., C. (2003). In Fore S. (Ed.), Reed canary grass control (displacement by a diversenative-species mix). KIRKSVILLE; 100 E NORMAL ST, KIRKSVILLE, MO 63501USA: TRUMAN STATE UNIV PRESS.Lavergne, S., & Molofsky, J. (2004). Reed canary grass (phalaris arundinacea) as abiological model in the study of plant invasions. Critical Reviews in Plant Sciences,23(5), 415-429. doi:10.1080/07352680490505934Lavergne, S., & Molofsky, J. (2006). Control strategies for the invasive reed canarygrass(phalaris arundinacea L) in north american wetlands: The need for an integratedmanagement plan. Natural Areas Journal, 26(2), 208-214.[208:CSFTIR]2.0.CO;2Lindig-Cisneros, R. & Zedler J.B. (2002). Phalaris arundinacea seedling establishment:effects of canopy complexity in fen, mesocosm, and restoration experiments. CanadianJournal of Botany, 80(6): 617-624Morrison, S. L., & Molofsky, J. (1998). Effects of genotypes, soil moisture, and competitionon the growth of an invasive grass, phalaris arundinacea (reed canary grass). CanadianJournal of Botany-Revue Canadienne De Botanique, 76(11), 1939-1946.doi:10.1139/b98-15732Perkins, L., & Nowak, R. (2012). Soil conditioning and plant—soil feedbacks affectcompetitive relationships between native and invasive grasses. Plant Ecology, 213(8),1337-1344.Reinhardt Adams, C., & Galatowitsch, S. M. (2006). Increasing the effectiveness of reedcanary grass (phalaris arundinacea L.) control in wet meadow restorations. RestorationEcology, 14(3), 441-451. doi:10.1111/j.1526-100X.2006.00152.xRust, N. A., Abrams, A., Challender, D. W. S., Chapron, G., Ghoddousi, A., Glikman, J. A.,. . . Hill, C. M. (2017). Quantity does not always mean quality: The importance ofqualitative social science in conservation research. Society & Natural Resources,30(10), 1304-1310. doi:10.1080/08941920.2017.1333661SPES (2013). Stanley Park Ecology Society Invasive Plant Management (2020). Stanley of the Park Report for the Ecological Integrity of Stanley Park., S. S., McEvoy, P. B., & Coombs, E. M. (2006). Negative per capita effects ofpurple loosestrife and reed canary grass on plant diversity of wetland communities.Diversity and Distributions, 12(4), 351-363. doi:10.1111/j.1366-9516.2006.00227.xSpyreas, G., Wilm, B. W., Plocher, A. E., Ketzner, D. M., Matthews, J. W., Ellis, J. L., et al.(2010). Biological consequences of invasion by reed canary grass (phalarisarundinacea). Biological Invasions, 12(5), 1253-1267. doi:10.1007/s10530-009-9544-yTu, M. (2004). Reed Canarygrass (Phalaris arundinacea L.) Control & Management in thePacific Northwest. The Nature Conservancy’s Wildland Invasive Species TeamWaggy, M.A. (2010). Phalaris arundinacea. Fire Effects Information System, [Online]. U.S.Department of Agriculture, Forest Service, Rocky Mountain Research Station, FireSciences Laboratory (Producer). Available: [2021, March 26].Watkins, D. C. (2012). Qualitative research: The importance of conducting research thatdoesn't "count". Health Promotion Practice, 13(2), 153-158.doi:10.1177/1524839912437370Weilhoefer, C. L., Williams, D., Nguyen, I., Jakstis, K., & Fischer, C. (2016). The effects ofreed canary grass (Phalaris arundinacea L.) on wetland habitat and arthropodcommunity composition in an urban freshwater wetland. Wetlands Ecology andManagement, 25(2), 159-175.33Woolf, N. H., & Silver, C. (2018). Qualitative analysis using NVivo the five level QDAmethod Routledge.Additional links,least%20an%20entire%20growing%20season. A1. Close-up view of certain regions of Stanley Park with high reed canary grass densities, shown with current SPES worksites overlaid in grey stripes.Locations clockwise from top left: Beaver Lake, Prospect Point, top right corner of Lost Lagoon, top left corner of Lost Lagoon.35Figure A2. Susceptibility raster for reed canary grass invasion/growth, considering light availability, undergrowth presence, hydrology, and trails. Lightercolours indicate higher susceptibility to reed canary grass invasion or growth.36Figure A3. Close-up view of certain regions of Stanley Park with high reed canary grass densities and their susceptibility. Lighter colours indicate highersusceptibility to reed canary grass invasion or growth. Locations clockwise from top left: Beaver Lake, Prospect Point, top right corner of Lost Lagoon, top leftcorner of Lost Lagoon.37Figure A4. Areas with high susceptibility grouped together using the ‘Locate Regions’ tool. Most risk areas (in red) have higher average cell value forsusceptibility. These areas were used to inform the drawing of the ‘Likely spread areas’ polygons.38Figure A5. Close-up view of regions of Stanley Park with high reed canary grass spread likelihood (Map 2), with likely spread areas shown in purple. Present reedcanary grass abundance shown in red (high density), orange (medium density), and yellow (low density). Locations clockwise from top left: Beaver Lake, ProspectPoint, Lost Lagoon, Seawall near stairs to Teahouse.39Figure A6. Close-up view of regions of Stanley Park with high reed canary grass spread likelihood (Map 2), with likely spread areas shown in purple, and SPEScurrent worksites shown in grey stripes. Present reed canary grass priority zones shown in red (high density), orange (medium density), and yellow (low density).Locations clockwise from top left: Beaver Lake, Prospect Point, Lost Lagoon, Seawall near stairs to Teahouse.40


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items