NWRS_InvasivePlantPilotProject_FinalReport_2014

National Wildlife Refuge System Invasive Plant Inventory Pilot Project Prepared by U S Fish and Wildlife Service National Wildlife Refuge System Inventory and Monitoring Initiative February 2014 USFWS Invasive Plant Inventory Pilot Project COVER PHOTO CREDITS (clockwise, upper left to lower left) Landscape view of San Diego National Wildlife Refuge U S Fish and Wildlife Service Map showing invasive plant occurrences at San Diego National Wildlife Refuge (Edvarchuk et al 2012) Alligator weed (Alternanthera philoxeroides) and phragmites (Phragmites spp.) invading a canal at Alligator River National Wildlife Refuge Photo by Corey Ransom, Utah State University Japanese stilt grass (Microstegium vimineum) invading a forest at Silvio O Conte National Fish and Wildlife Refuge Photo by Kim Edvarchuk, Utah State University Predicted distribution of Alligator weed in the South Atlantic Landscape Conservation Cooperative using the Maxent model (Evangelista et al 2012) Landscape view of Quivira National Wildlife Refuge Photo by Kim Edvarchuk, Utah State University Suggested Citation: U S Fish and Wildlife Service 2014 National Wildlife Refuge System invasive plant inventory pilot project National Wildlife Refuge System, Inventory and Monitoring Initiative, Fort Collins, USA USFWS Invasive Plant Inventory Pilot Project PROJECT TEAM *Jenny Ericson, USFWS, Region 9, National Invasive Species Coordinator (project lead) *Lindy Garner, USFWS, Region 6, Invasive Species and Integrated Pest Management Coordinator *Giselle Block, USFWS, Region 8, Inventory and Monitoring Specialist *David Bishop, USFWS, Region 4, Invasive Species Coordinator *Kim Edvarchuk, Utah State University, Department of Plants, Soils and Climate (lead: field inventories) *Paul Evangelista, Colorado State University, Natural Resource Ecology Laboratory (lead: modeling) Thomas Stohlgren, USGS, Fort Collins Research Center Andres Abeyta, IGIS Technologies, Inc (lead: RLGIS training) *Report authors ACKNOWLEDGEMENTS Staff from Alligator River National Wildlife Refuge (NWR), Quivira NWR, Silvio O Conte National Fish and Wildlife Refuge (NFWR) and San Diego NWR made significant contributions to the Invasive Plant Inventory Pilot Project; without them the project would not have been possible In particular we thank the following refuge staff: • • • • San Diego NWR: Andy Yuen, Project Leader; Slader Buck, Deputy Project Leader; Jill Terp, Refuge Manager; John Martin, Refuge Biologist; Pek Pum, GIS Specialist Alligator River NWR: Mike Bryant, Project Leader; Scott Lanier, Deputy Project Leader; Brian Van Druten, Refuge Biologist; Dennis Stewart, Refuge Biologist Quivira NWR: Dan Severson, Refuge Manager; Rachel Laubhan, Refuge Biologist Silvio O Conte NFWR: Andrew French, Project Leader; Cynthia Boettner, Invasive Plant Control Initiative Coordinator We thank Mark Chase and Jana Newman for their continued support of projects aimed at improving invasive species management in the National Wildlife Refuge System We thank Sean Blomquist (R3 Zone Biologist), Kevin Kilbride (R1 I&M Coordinator), Bridgette Flanders-Wanner (R1 Regional Biologist), Orien Richmond (R8 I&M Biologist) and Sharon Dulava (R8 I&M Biologist) who provided helpful comments on this report Special thanks to Erin Boyle (R8 GIS Specialist) for her assistance with map development Lastly, a special thanks to refuge partners who contributed their time and Jenny Ericson was acting National Invasive Species Coordinator until early 2013 She is currently the Deputy Field Office Supervisor at the USFWS Yreka Fish and Wildlife Office in northern California USFWS Invasive Plant Inventory Pilot Project expertise; their insights and knowledge greatly enhanced this project In particular, we thank the U.S Air Force, The Nature Conservancy, AECOM Technology Corporation, Sweetwater Authority, California Department of Fish and Wildlife, RECON Environmental, Inc and the Conservation Biology Institute Funding for this project was provided by the National Wildlife Refuge System Inventory and Monitoring Initiative Table of Contents EXECUTIVE SUMMARY INTRODUCTION 11 2.1 Background, Need and Purpose 11 2.2 Project Goal, Objectives and Activities 12 METHODS 14 3.1 Phase I: Selection of Refuge Pilot Sites 15 3.1.1 Pilot Refuge Background Information 18 3.2 Phase II: Inventory Planning Workshops .24 3.2.1 Inventory Field Methods 24 3.3 Phase III: Data Collection and Management Training 26 3.4 Phase IV: Invasive Plant Inventory 26 3.5 Phase V: Inventory Reporting and Species Distribution Modeling 26 3.5.1 Inventory Reports and Data Sharing .26 3.5.2 Ecological Niche Modeling .27 3.6 Phase VI: Refuge Follow-up 29 RESULTS .30 4.1 Inventory Planning 30 4.1.1 Inventory Objectives .30 4.1.2 Inventory Target Species 33 4.1.3 Inventory Target Areas 36 4.2 Refuge Inventory Field Methods 39 4.3 Refuge Inventory 41 4.4 Modeling 45 4.5 Refuge Follow-up 48 DISCUSSION 49 5.1 Workshops 50 5.2 Inventory Objectives 50 USFWS Invasive Plant Inventory Pilot Project 5.3 Species Prioritization 51 5.4 Area Prioritization 53 5.5 Inventory Methods and Results 53 5.6 Data Collection and Management 55 5.7 Habitat Suitability Modeling 56 6.0 Recommendations 57 Bibliography .59 Tables Table National Wildlife Refuge System Invasive Plant Inventory Pilot Project: characteristics of participating National Wildlife Refuges (N=4) 16 Table National Wildlife Refuge System Invasive Plant Inventory Pilot Project (2011-2013): inventory data utility across four refuges 31 Table National Wildlife Refuge System Invasive Plant Inventory Pilot Project (2011-2013): invasive plant inventory objectives across four refuges .31 Table National Wildlife Refuge System Invasive Plant Inventory Pilot Project (2011-2013): criteria used to select and prioritize species targets across four refuges 35 Table National Wildlife Refuge System Invasive Plant Inventory Pilot Project (2011-2013): criteria used to select inventory areas across four refuges 38 Table National Wildlife Refuge System Invasive Plant Inventory Pilot Project (2011-2013): inventory data collection parameters across four refuges 40 Table National Wildlife Refuge System Invasive Plant Inventory Pilot Project (2011-2013): invasive plant inventory results across four refuges 42 Table Dominant environmental predictors of future invasive plant species distributions at refuge (N=4) and associated Landscape Conservation Cooperative scales .46 Table Comparison of invasive plant distribution modeling results across four National Wildlife Refuges, 20112012 .47 Figures Figure Workflow of the National Wildlife Refuge System Invasive Plant Inventory Pilot Project, 2011-2013 14 Figure Location of refuges participating in the National Wildlife Refuge System Invasive Plant Inventory Pilot Project: 2011-2012 17 Figure Alligator River National Wildlife Refuge and surrounding lands .20 Figure Silvio Conte National Fish and Wildlife Refuge and surrounding lands .21 Figure Quivira National Wildlife Refuge and surrounding lands 22 Figure San Diego National Wildlife Refuge and surrounding lands .23 Figure Invasive plant inventory effort and level of infestation across four National Wildlife Refuges: 20112012 .43 Figure The invasion s-curve (adapted from Department of Environment and Primary Industries 2010) .52 Figure 9.Relationship between eradication success, size of infestation and management effort … …………….55 USFWS Invasive Plant Inventory Pilot Project Appendices Appendix A Invasive Plant Inventory Planning Questionnaire Appendix B Invasive Plant Inventory Report: Alligator River National Wildlife Refuge (2012) Appendix C Invasive Plant Inventory Report: Quivira National Wildlife Refuge (2012) Appendix D Invasive Plant Inventory Report: Silvio O Conte National Fish and Wildlife Refuge (2012) Appendix E Invasive Plant Inventory Report: San Diego National Wildlife Refuge (2012) Appendix F Mapping Habitat and Potential Distributions of Invasive Plant Species on USFWS National Wildlife Refuges (2012) EXECUTIVE SUMMARY Invasive species are one of the most pervasive threats to biological integrity, diversity, and environmental health on units of the National Wildlife Refuge System (NWRS) Invasive species can alter ecosystem function and structure (e.g., shift fire regimes or nutrient cycling) and change the distribution, abundance and diversity of native species and communities The high cost of managing invasive species combined with decreasing budgets requires that we prioritize which invasive threats to manage and we use scientifically credible data to inform optimal management decisions and evaluate progress A key first step in invasive management is identifying the species of greatest management concern and conducting an inventory of those species to determine their distribution and abundance Inventory data inform management decisions and provide a baseline for evaluating progress Although the value of inventory data is well known, inventories are considered a daunting task for many resource managers because of limited resources or lack of expertise on how to conduct invasive inventories To begin addressing these challenges, the NWRS Inventory and Monitoring Initiative and the NWRS Invasive Species Program initiated the NWRS Invasive Plant Pilot Project (Invasive Pilot) to examine the diversity of factors influencing how refuges plan and conduct invasive plant inventories The goals of the Project are to: (1) evaluate how variation in refuge size, environment, resource availability and other factors influence planning and implementation of invasive plant inventories and (2) use this information to inform a NWRS framework or guide for conducting invasive plant inventories Although the focus of this effort is on invasive plants, many of the concepts can be applied to other invasive taxa This project was administered by invasive plant subject matter experts employed by the U S Fish and Wildlife Service, Utah State University, and Colorado State University Invasive Plant Inventory Planning The Invasive Pilot was conducted at four National Wildlife Refuges (NWR) between 2011 and 2012: Silvio O Conte National Fish and Wildlife Refuge (Region 5), Alligator River NWR (Region 4), Quivira NWR (Region 6) and San Diego NWR (Region 8) Refuge size ranged from 9,235 acres to over 153,000 acres Refuge environments spanned prairie grasslands, coastal wetlands, broadleaf forest and coastal sage scrub/chaparral Refuge settings ranged from rural to highly urban (surrounded by >3million people) All of the refuges in the Invasive Pilot were actively managing invasive plants and recognized the benefits of inventory data Despite this, little quantitative abundance or distribution data had been collected for their highest-ranked invasive threats and, where data was collected, standardized protocols USFWS Invasive Plant Inventory Pilot Project were lacking Common reasons for not conducting inventories include a lack of staff and funding, lack of knowledge about how to plan and conduct invasive plant inventories or a belief that inventories are not an efficient use of resources (relative to treatment) Invasive Pilot activities at each refuge included a 2-day inventory planning workshop, a 2-day data collection and management training, field-based invasive plant inventories (2-4 weeks) and species distribution modeling Although a common inventory planning process was applied across refuges, there were differences in inventory objectives, dominant factors used to select inventory targets (species, areas), field methods, results (e.g., acres inventoried per unit time) and near-term data use This variability was driven by many factors including the number and magnitude of invasive plant threats, invasive management history, knowledge about invasive ecology and management, refuge size and accessibility, vegetation types and diversity, partner engagement, presence of other stressors or sensitive resources and management resource limitations At all refuges, informing strategic management of invasive plant species in order to reduce stress on priority natural resources was a fundamental reason for conducting an inventory Refuges documented one or more ways they intended to use the data Data utility fell into four major categories: (1) inform conservation plans and strategies (e.g., Integrated Pest Management Plan, prioritize/direct control actions, estimate control costs, design control efficacy monitoring); (2) detect and eradicate new and highly invasive plant species; (3) support landscape-level partnerships and awareness and (4) increase ecological understanding (e.g., identify pathways of spread) Refuge invasive plant inventory objectives fell into four categories: (1) identify invasion boundaries and satellite populations; (2) identify locations of new invasive plant species; (3) identify invasive-free prevention areas and (4) document distribution and abundance of established species Inventory objectives were further refined by the selection of target species and areas Inventory target species and areas were refined and, where needed, prioritized during inventory planning workshops using refuge knowledge, partner knowledge and other sources of external data (e.g., regional invasive plant databases) The number of inventory target species was at Alligator River, 17 at Quivira, 16 at Silvio O Conte and 24 at San Diego A total of ten criteria were developed by refuges to select inventory target species (not all criteria were used by each refuge) The criteria included inventory objectives, species invasiveness, species impacts to wildlands (current, potential), species’ management priority, belief about species abundance/distribution and species detectability The need for species prioritization was greatest at San Diego and Silvio O Conte and both refuges were in close proximity to highly urbanized areas Refuges in close proximity to urban areas or with long history of human disturbance are more likely to have a need for species prioritization Pilot refuges also prioritized areas for inventory because it was unlikely that all refuge lands would be inventoried given available mapping resources A total of fifteen criteria were developed by refuges to prioritize areas for inventory (not all criteria were used by each refuge) Criteria common to all refuges included: (1) presence or diversity of a landcover type, vegetation community or ecosystem; areas with a diverse, unique or priority landcover type, vegetation community or ecosystem were given a higher priority; and (2) management investment history; areas where resource investments (e.g., restoration, invasive plant management) were planned or had recently occurred were given a higher priority USFWS Invasive Plant Inventory Pilot Project Major Recommendations: Develop a NWRS guide or framework for invasive plant inventory and early detection that includes (1) guidance on planning invasive plant inventories and early detection; the inventory planning process promotes a shared acceptance and understanding of objectives, priorities, methods and outcomes; promotes collaboration and promotes efficient and effective use of refuge resources; (2) Develop tools for selecting and prioritizing refuge-specific invasive plant threats, including the use of external information (e.g., experts, invasive databases, scientific literature); (3) provide field-level refuge support for use of the framework and associated toolbox Field Inventory Methods Field inventory methods were based on the Utah State University Wildland Weed Mapping Methods (USU method) and modified to meet refuge-specific inventory objectives Data collection standards were similar across refuges except for minimum detection target size, effective detection swath width and patch separation resolution Factors influencing refuge-specific data collection standards and search patterns included the density and structure of vegetation, target species visibility, the presence of sensitive species (e.g., federally endangered species) and the level of spatial resolution desired by refuge staff Major Recommendations: Develop a NWRS framework for invasive plant early detection and inventory that: (1) identifies and describes the range of existing inventory and early detection methods; (2) provides inventory and early detection framework protocols; (3) provides for flexibility in determining methods conditional on site-specific factors (e.g., objectives, scale); (4) identifies appropriate conditions for protocol-specific use; (5) identifies protocol-specific resource and technical expertise requirements; (6) provides options for data management structures and data management guidance; and (7) provides technical support at the zone or regional level to assist refuges with database development and data management issues Field Inventory Results The USU five-person mapping team conducted inventories for a period of two weeks at Alligator River, Quivira and Silvio O Conte At San Diego, the inventory was conducted for weeks The number of acres inventoried was highest at Quivira (10,163ac; 46% of refuge lands), followed by Alligator River (8,989ac; 6% of refuge lands), San Diego (1,962ac; 21% of refuge lands), and Silvio O Conte (1,358ac; % of refuge lands) The percentage of refuge lands inventoried was influenced by refuge size and the average number of acres a surveyor could search within a day Factors influencing number of acres inventoried per unit effort were accessibility, vegetation density and structure, number of species targets and abundance of target species Dense forest vegetation at Silvio O Conte NWR and dense marsh vegetation at Alligator River required more search effort per unit time relative to open environments such as grasslands at Quivira At San Diego NWR, the large number of invasive species targets (N=24) and the high abundance of some species (e.g., non-native annual grasses) reduced the number of acres that could be searched per unit time relative to the other pilot refuges Although useful data was collected at each refuge, it is clear that inventory methods not only must be suited to survey objectives but must take into account refuge size, environment, accessibility and USFWS Invasive Plant Inventory Pilot Project available resources A number of invasive plant inventory methods have been developed by conservation organizations that differ in the type of inventory objectives that can be satisfied and the time and equipment requirements Because of the variability in refuge management objectives, environments, spatial scope and available resources, no single method for assessing invasive plant distribution and abundance can meet all refuge needs Percent of inventory area infested by target species was similar among of the refuges: Alligator River (2.2%), Silvio O Conte (1.9%) and Quivira (1.3%) In contrast, 58% of the inventory area was infested with target species at San Diego Percent of total infested acres represented by targets designated by refuge personnel as early detection species was similar across refuges and ranged from 0.2 (Quivira, San Diego) to 2.2 (Alligator) (mean=0.9%) This result shows that refuges were aware of certain species that are in the early stages of invasion and demonstrates how early detection monitoring can support eradication of new species and thus avoid future costs and ecological damage Major Recommendations: (1) develop regional or national mechanisms that will encourage and support early detection and rapid response of invasive species in order to prevent future ecological and economic harm; (2) see previous Methods recommendations Although a NWRS framework for invasive plant early detection and inventory will provide valuable planning and technical guidance, additional refuge-level support is needed Species Distribution Modeling Species distribution models (also known as ecological niche models or habitat suitability models) were developed for two invasive plant species at each pilot refuge Models were developed at the refuge scale and at the corresponding Landscape Conservation Cooperative (LCC) scale Maxent models were used to demonstrate to refuges how inventory or early detection information can be used to forecast invasive species distributions The resulting models highlight environmental conditions that facilitate future spread and inform priority locations for prevention, monitoring, and management Biotic and abiotic predictor variables and refuge-specific species occurrence data were used to construct refuge-specific species distribution models Common predictor variables at the refuge scale were aspect, distance to water, elevation, slope, soils, and land cover type The quality and quantity of refuge-generated datasets varied considerably and additional external data sources were used (e.g, elevation, hydrology), when needed, to support modeling efforts For model runs at the LCC scale, 32 environmental predictor variables were used These included 19 seasonal climatic indices and variables such as elevation, flow accumulation, distance from water, slope (i.e., degrees), geology and solar radiation All models suggested that selected invasive plant species have the potential to expand beyond current boundaries at both the refuge and larger LCC scale Because many of the target species selected for modeling are associated with water features either as a habitat feature or vector, areas most at risk were those associated with canals, river drainages, distance to water, and other water sources For San Diego and the California LCC, soils, land cover, slope, and bioclimatic variables were the top contributing predictors In this case, the models USFWS Invasive Plant Inventory Pilot Project helped highlight locations of likely infestation that would not be intuitively tied to a particular feature such as a waterway or other vector pathway Major recommendations: (1) provide regional support to identify, evaluate, and increase refuge-level access to environmental (abiotic) datasets that are commonly used to model invasive plant species distributions; (2) support the development of training (e.g., NCTC) for regional and refuge staff on how to use software such as Maxent to model potential invasive species distributions; and (3) incorporate species distribution modeling methods and utility in the future NWRS framework for invasive plant early detection and inventory Refuge Follow-up Follow-up interviews were conducted with refuge staff to assess the usefulness of the inventory planning process, GPS/RLGIS training and inventory and modeling results to inform invasive plant management All pilot refuges said the inventory planning process was very useful The planning elements considered most useful were prioritization of species and areas and facilitation by invasive experts Refuge inventory results have already been used by refuges to direct treatments, obtain treatment funding, identify invasive “hot-spots” and inform management planning (CCPs, HMPs and IPMs) Three of the four refuges intend to expand or repeat invasive plant inventories in future years Quivira and San Diego plan to use the USU Mapping Method for early detection along invasive pathways such as roads and trails These refuges, along with Silvio O Conte, are also interested in exploring other less-intensive field methods (e.g., grid-based methods) to increase the number of acres surveyed per unit effort These less-intensive methods would be used to compliment early detection surveys along vector pathways (e.g., roads, trails, waterways) Key Words: early detection, invasive species, inventory, monitoring, National Wildlife Refuge System, plants, species distribution model, stressors, threats 10 4.5 Refuge Follow-up Follow-up interviews were conducted with refuge staff to assess the usefulness of the inventory planning process, GPS/RLGIS training and inventory and modeling results to inform invasive plant management These interviews were conducted at least months after receipt of the inventory and modeling reports and associated data Inventory planning workshop All pilot refuges said the inventory planning process was very useful The planning elements considered most useful were prioritization of species and areas, facilitation by invasive experts, identification of data that is critical to collect during an inventory and gaining a better understanding of the many uses of inventory data Refuges said the process of prioritizing target species highlighted factors they would consider in the future when developing invasive management plans and deciding how to allocate limited resources San Diego has already used the prioritization results to inform development of an Integrated Pest Management Plan, and three of the four refuges said lessons learned from the workshop will inform planning processes in the future (e.g., IPM plans) Inventory results Refuge inventory results have already been used by refuges to direct treatments, obtain treatment funding, identify invasive “hot-spots” and inform management planning (CCP, HMP, IPM) All of the pilot refuges are using the inventory data to direct eradication of early detection species or satellite populations of established species The data is also being used is to adjust spatial treatment strategies For example, refuges are refocusing treatment resources on outlying invasive plant populations rather than large core infestations The data has also been used by at least two refuges to direct removal of early detection species by volunteers Inventory methods Three of the four refuges intend to expand or repeat invasive plant inventories in future years Quivira and San Diego plan to use the USU Mapping Method for early detection along invasive pathways such as roads and trails These refuges, along with Silvio O Conte, are also interested in exploring other less-intensive methods to increase acres surveyed per unit effort These less-intensive methods would be used to compliment early detection surveys along vector pathways (e.g., roads, trails, waterways) Alligator does not intend to use the USU invasive inventory method because of limited resources and will instead record invasive plant locations concomitantly with invasive plant treatments Alligator is considering the use of remote methods (e.g., aerial imagery) to survey highly visible species such as phragmites in densely vegetated coastal marsh RLGIS and databases Quivira will continue the use of RLGIS to collect and store inventory data Silvio O Conte and San Diego said that they are not likely to use RLGIS in the future because the system is not supported in their region and/or it was considered a cumbersome data collection system Despite this, the RLGIS invasive plant feature class dictionaries were considered useful and will likely be integrated into other spatial data collection systems Alligator may continue using RLGIS but stated its use may be limited due to issues with software compatibility Silvio O Conte and San Diego both expressed the need for database structures that can integrate inventory data with treatment and monitoring information through time 48 USFWS Invasive Plant Inventory Pilot Project Modeling results None of the refuges had thoroughly read the species distribution modeling report and could therefore not comment its utility Despite this, Quivira has successfully used some of the modeling results, in conjunction with inventory results, to obtain grant funding for invasive plant treatments DISCUSSION Results from the NWRS invasive plant inventory pilot project (Project) provide examples of inventory planning, execution, results and near-term utility across a diverse set of refuges Although a common inventory planning process was applied across the four refuges, there were differences in inventory objectives, criteria used to select inventory targets (species, areas), field methods, results (e.g., acres inventoried per unit time) and near-term data use This variability was driven by many factors including the number and magnitude of invasive plant threats, invasive management history, knowledge about invasive ecology and management, refuge size and accessibility, vegetation types and diversity, partner engagement, presence of other stressors or sensitive resources and management resource limitations This Project demonstrates the many challenges refuges face in planning and carrying out invasive plant inventories Planning and conducting invasive plant inventories requires knowledge about invasive plant ecology, survey design, spatial data collection and data management systems Few refuges have staff with this breadth of expertise or the resources to invest in meeting these needs All of the refuges in this study were actively managing invasive plants and recognized the benefits of inventory data Despite this, little quantitative abundance or distribution data had been collected for their highest-ranked invasive threats, and therefore SMART threat reduction objectives were absent Common reasons for not conducting inventories included a lack of staff and funding, lack of knowledge about how to plan and conduct invasive plant inventories or a belief that inventories are not an efficient use of resources (relative to treatment) One of the major challenges of the NWRS is shifting the treatment-dominated invasive management approach to an approach where prioritization (species, areas) and collection of information (inventory, monitoring) are considered essential parts of invasive plant management The NWRS Inventory and Monitoring Initiative is in a unique position to promote and support this shift by: • Providing support for: (1) invasive management approaches that incorporate learning and adapting; (2) development of a framework and site-specific invasive inventory and monitoring protocols; (3) development of invasive management prioritization tools; (4) database development; and (5) data analysis • Demonstrating the utility of inventory and monitoring data to improve invasive plant management outcomes using available case studies Without baseline invasive species distribution and abundance data, conservation land managers are unable to develop measurable threat-reduction objectives, optimally allocate management resources or measure progress Over the long term this lack of baseline information can lead to low economic and ecological returns Developing a framework or guide to invasive plant inventories is one of several steps that can be taken to support management of invasives in the NWRS Such a framework should provide guidance on how to plan invasive plant inventories including how to develop survey objectives, prioritize invasive species, identify optimal survey methods and manage data In addition, field-level 49 USFWS Invasive Plant Inventory Pilot Project support is needed to develop site-specific inventory protocols and data management systems that meet refuge needs, available resources and level of expertise 5.1 Workshops The workshop process was crucial for refuge staff and their partners to share everyone’s beliefs about invasive plant issues and expectations for the inventory There are often unspecified beliefs or expectations that differ among individual refuge personnel (and partners), and these differences should be resolved before investing valuable resources The inventory planning process promotes a shared acceptance and understanding of objectives, priorities, methods and outcomes, promotes collaboration and promotes efficient and effective use of refuge resources The workshops also provided a unique opportunity for refuge staff to think critically about how invasive plants impact priority conservation targets and impair achievement of refuge conservation goals In addition, refuges were able to discuss and rank invasive plant threats, review how threats are currently managed and how inventory information could improve management For example, knowing invasive plant abundance and distribution information allows development of measurable threat-reduction objectives (e.g., reduce cover of species A by 25% in years, eradicate satellite populations of species B in area X within years) An important benefit of inventories is the documentation of outlying or satellite populations of priority invasive species This information can inform containment strategies aimed at preventing future spread of a priority invasive species This strategy is especially useful when resources are limiting Refuges are typically aware of large infestations but may be less knowledgeable about the location of satellite populations or invasion boundaries Partner engagement was a critical component of the inventory planning process By engaging partners, pilot refuges were provided with a larger landscape view of invasive plant threats, invasive impacts to natural resources, current management practices and ecological knowledge Partner input influenced inventory planning decisions at most of the pilot refuges, especially with respect to prioritizing invasive species targets Engagement with partners also promoted new opportunities for invasive species management partnerships 5.2 Inventory Objectives Refuge inventory (survey) objectives ranged from identifying invasion boundaries of established species to identifying locations of new highly invasive species Because survey objectives drive decisions about survey design (e.g., detection confidence, patch separation resolution) and resource needs, refuges were encouraged to develop SMART objectives: Specific, Measurable, Achievable, Realistic and Time scaled An objective that is SMART is more likely to be achieved because it is clear (specific), you can tell when it has been achieved (measurable), it’s an action that is feasible and you know when it should be achieved Inventory objectives should identify the target (species, area), identify when the inventory will occur and be realistic about what can be accomplished given available resources Refuges were also encouraged to document the intended use of the data Pilot refuges often listed more than one objective and cited multiple uses of the data In some cases this can create issues because the selection of target species and areas, methods and scale of data collection are linked to survey objectives, and when an inventory tries to address multiple survey objectives it reduces the likelihood of achieving any single objective For example, methods designed for documenting the distribution and abundance of 50 USFWS Invasive Plant Inventory Pilot Project established species may not work well for detecting rare species and vice versa The planning workshops worked very well to illustrate this point and helped encourage staff to develop objectives that were realistic and achievable and to be explicit about how the data would be used to inform management Inventory data can be used to inform or support invasive plant management in a number of important ways: • Identify locations of uncommon and highly invasive species for eradication (early detection and rapid response) • Identify and maintain invasive plant-free areas (clean areas) • Identify perimeters and satellite populations of established invasive species for containment • Estimate treatment costs • Inform spatial treatment strategies, espcially given limited and variable management resources (staff, funds) • Identify invasive “hot-spots” and patterns of spread to guide prevention activities and future monitoring • Assess impacts to sensitive resources or relationships to other environmental variables • Demonstrate invasive issues to refuge leadership and conservation partners • Support landscape-level invasive management • Increase capacity to acquire funding • Inform management effectiveness monitoring design Higher level refuge conservation goals, objectives and strategies were discussed during development of inventory objectives and influenced what species were targeted and where inventories should occur Often, refuge management objectives identified the need to manage particular invasive threats in particular management units Refuges also described threat-reduction strategies, which focused primarily on treatment types or locations What was lacking in all cases were SMART threat-reduction objectives (e.g “within unit A, reduce species X by 20% within years”), and this is likely due to the absence of baseline abundance and distribution information Inventories provide the means for developing SMART threat reduction strategies and form the basis for learning and evaluating progress 5.3 Species Prioritization Because resources are often limiting and not all non-native plants have equivalent impacts on native species and ecosystems, it is not feasible (or necessary) to inventory all non-native plants in a given area A necessary first step is to identify and document which non-native plant species cause harm or have the potential to cause harm (i.e., are invasive) to refuge natural resources of conservation concern This information can be derived from refuge knowledge and augmented with external information on species ecology and known impacts to wildlands in environments similar to the refuge This information is available from a variety of sources including state noxious weed lists, invasive species councils, early detection networks, local weed management areas and published literature Where the number of invasive species (current or potential) exceeds what can be feasibly inventoried or managed, prioritization is essential The need for species prioritization is expected to vary as shown by this 51 USFWS Invasive Plant Inventory Pilot Project Project Urban refuges, such as San Diego and Silvio O Conte, with a long history of human disturbance are more likely to have an extensive list of invasive species and therefore a greater need for prioritization In all cases, it is essential that refuges regularly assess and document which invasive species should be a focus of management and why Such information provides a foundation for optimal allocation of management resources over time The Project highlighted several criteria that should be considered when prioritizing species for inventory or management The level of observed harm or potential harm to trust resources and invasion stage were common criteria influencing species selection Once introduced, invasive species often follow a trajectory depicted by the “invasive s-curve” (Figure 8) Recognizing which species are most harmful to conservation targets (currently or in the future) and identifying the respective stage of invasion are important factors to consider when deciding how to allocate limited resources for inventory or management (e.g., prevention, early detection/surveillance, containment) In addition to mining refuge information, engaging conservation partners and reviewing external invasive plant information (e.g CA Invasive Plant Inventory, EDDMapS, Invasive.org) are steps that should be taken when prioritizing species for inventory and management External sources of information can highlight locations of new and harmful threats, identify regional species trends and provide information on species ecology and impacts Figure The invasion s-curve (adapted from Department of Environment and Primary Industries 2010) 52 USFWS Invasive Plant Inventory Pilot Project 5.4 Area Prioritization This project highlighted several criteria that should be considered when prioritizing areas for inventory and ultimately management As with selection of target species, higher-level refuge management goals and objectives should have a strong influence on decisions about what to survey and where to conduct surveys For example, if management is focused on conservation of endangered species or a particular ecosystem, areas where these resources occur may be a strongly weighted factor in the prioritization process Management objectives focused on early detection may strongly weight areas with a high density of potential spread pathways (e.g., roads, trails), whereas objectives focused on a single invasive plant species may instead focus on areas with suitable habitat conditions Similar to species prioritization, each of the four pilot refuges developed a unique set of criteria to select inventory areas suggesting flexibility is needed as to which criteria are used or how to weight the importance of individual criteria Prioritization of inventory areas proved more difficult than selecting target species due to to the variety of environments, complexity of conservation objectives and the realization that all areas could not be inventoried within the time frame of the Project Because the total area that can be inventoried within a growing season will be constrained by available resources, phasing of inventories in space and time will be needed at many refuges Although locations of priority conservation resources (species, communities, ecosystems) was a dominant criteria used by refuges to prioritize inventory areas, several other types of information was used including: • Refuge management unit boundaries • Unit size • Waterways, roads, trails, levees and other potential invasive spread pathways • Vegetation communities or landcover types • Areas of recent management investment (e.g., restoration) • Disturbance history (e.g., fire, floods) • Soil types • Topography • Perceived level of invasion by invasive plants 5.5 Inventory Methods and Results Documenting the spatial distribution and abundance of invasive plants is a daunting task for most natural resource managers Total acres or the percent of refuge land inventoried during this project varied across refuges and was influenced primarily by refuge size, vegetation structure and accessibility This project demonstrates that inventory methods not only must be suited to survey objectives but must take into account refuge size, environment, accessibility and available resources A number of invasive plant inventory methods have been developed by conservation organizations that differ in the type of inventory objectives that can be satisfied and the time and equipment requirements Because of the variability in refuge management objectives, environments, spatial scope and available resources, no single method for assessing invasive plant distribution and abundance can meet all refuge needs A future NWRS framework for invasive plant inventories should identify the range of existing inventory and early detection methods, their optimal conditions for use and resource and expertise requirements 53 USFWS Invasive Plant Inventory Pilot Project Invasive plant inventories for this project were conducted using the Utah State University Wildland Weed Mapping method (USU method) This method has been applied throughout the United States, including on National Park Service and U.S Forest Service lands The USU method is rigorous in that it requires the user to define several standards for field data collection and accuracy, but it is a more laborintensive method because it involves complete searches of a defined area and documentation of both location and size of individual invasive plant occurrences This method is especially useful for detection of species that are uncommon or rare (early detection) or where the objective is to detect outlying populations of established species The USU inventory method is less appropriate for inventory of widely distributed and abundant species, where visibility is low (e.g., dense vegetation) or where the inventory area is large and must be covered in a short period of time Vegetation characteristics greatly influenced total acres surveyed per unit effort across pilot refuges For example, surveys in dense forested areas of Silvio O Conte resulted in fewer acres searched per unit time relative to the other pilot refuges At San Diego, inventory of target species that were well established reduced the acres inventoried per unit time To reduce inventory effort, estimates of patch were recorded using radius (points) and width (lines) rather than walking entire patch perimeters Recent research suggests there is no significant difference in patch size accuracy between a buffered-point, screen-drawn polygon, and a perimeter-walked polygon (Christensen et al 2011) Percent infestation of species labeled as ‘early detection species’ by refuge personnel prior to the inventory was low (