Supporting Research on Invasive Species and their Control

The gathering of critical ecological data that characterizes the biology, distribution, and impacts of three invasive species, the biology and ecology of potential biological control agents at population and community levels, and the development of new re-vegetation technologies for area-wide implementation.

Overview: In Task 2, detailed experimental activities will be conducted to provide the underlying scientific knowledge that will allow the effective implementation of area-wide management of the target invasive weeds. Through previous and on-going projects, preliminary work has already been accomplished that includes, the development of life-table data for YST growth and development (DiTomaso et al. 1998); the reproductive biology of Arundo donax (Holt and Boose, 1998); genetic characterization of invasive plant populations including saltcedar (Gaskin 1998); the introduction of six natural enemies for YST (Pitcairn et al. 1998); and extensive identification and host-specificity testing of natural enemies for saltcedar (Deloach 1987). Additional research is also available on a variety of nonbiologically-based management tactics (see DiTomaso 1999) and various habitat restoration methods to prevent the invasion of additional exotic species (Allen 1998). However, much experimentation is still required to develop an adequate understanding of the organismal biology and their ecosystem level dynamics, methods to gain control over appropriate management tactics for these exotic and invasive weed problems on a landscape level, and for integrating biologically-based technologies into an overall management scheme. It is especially true that biologically-based technologies have not been well integrated with other management activities and actually may be in conflict with other efforts conducted for invasive weed control.

To accomplish our overall project goal and the responsibilities of Task 2, we have developed 5 objectives that will be investigated by three interacting research teams centered in CA/NV, CO/WY, and NM/TX. These groups will conduct integrated studies on the listed objectives, some independently and others in complete cooperation across areas. These studies will be conducted at or in association with the 13 study sites discussed in more detail in Task 3 (see figures 4 and 5), however, the majority of the in depth research activities will be conducted on one of three more fully instrumented and biologically characterized study locations. Due to limited space, only one of these in depth research locations will be expanded upon in the proposal but realize that two additional sites in the other regions are also available and will be used as primary locations of investigation. The riparian drainage of Cache Creek (figure 5) is a highly diverse riverine system that supplies water and other resources to agricultural, urban and natural systems that surround the area. The Cache Creek drainage is highly infested with all three of the target invasive weeds that are the targets of our investigation. We plan to work along a 30 mile section of Cache Creek from BLM land in the higher elevations near Clear Lake, CA (ca. 4000 ft.) to the City of Woodland, CA where the Cache Creek settling basin connects into the Sacramento River drainage (ca. 50 ft. elevation). In all research locations, detailed aerial imagery has been collected to characterize both surrounding land use categories and vegetation types. Extensive sets of historical photography are available all along Cache Creek for each of the past ten years or more. This imagery is of high quality allowing the identification and mapping down to the detail of individual shrubs (see the red colored Tamarix parviflora shown in high magnification photographs of figure 5). Detailed habitat and vegetation maps will be developed for each of these sites using GIS/GPS technology and used as a base for both specific experimental studies and overall landscape level evaluations. Watershed areas will not only be characterized through remote sensing but will be instrumented throughout to collect adequate weather and physical data and ground sampled to verify vegetation type and density. These base maps will be used for a variety of experimental and predictive purposes over the duration of the study period. Within each experimental area, several critical landowners and land managers have been contacted and plan to participate in our research and implementation activities. Along Cache Creek for example, we have arranged to work with several private landowners, the Bureau of Land Management, the Cache Creek Conservancy, the Wintun Native American Council, the Yolo County Flood Control District and the Town of Woodland, CA (see attached example letters of support, Appendix B). Other supporting groups such as The Nature Conservancy, Team Arundo del Norte, California Department of Fish and Game, Natural Resource Conservation Service and the US FWS, and others support this effort.

Similar infrastructure and supporting groups are associated with our other two detailed research sites near Pueblo, CO and Seymore, TX. As the project progresses, a significant amount of the detailed research will be transitioned to more applied activities and conducted at one or more of the 13 implementation sites discussed in Task 3. All 13 implementation sites have also been documented using aerial photography (by Consortium member G. Anderson) to provide baseline habitat information and will be ground sampled during the first year of this project. In addition to work being conducted at our detailed worksites, some of our specific experimental activities such as revegetation methodologies will be conducted at other locations where specific local attributes are critical for the experiments of concern. The following objectives highlight the primary focus of Task 2, however, due to space limitations not all studies or the associated methods could be fully described here, therefore the following represents only a subsample of what has been planned.

Objective 2a: Develop a qualitative and quantitative understanding of invasive plant growth, development, reproductive biology and competitive relations that make YST, saltcedar and giant reed, highly invasive and destructive to US agriculture and the environment.

Justification: As a basis for all our investigations, plant growth, development and reproductive biology will be investigated for saltcedar and Arundo donax under differing abiotic and biotic situations characteristic of the growing areas where they are abundant and invasive. Similar data for YST is already available through Consortium partners. We hypothesize that these data will be useful to assess plant growth potential in study areas and factors such as biological control agents that may inhibit their growth. Plant growth models (both structural (3-d) and physiologically-based, will be developed and used in comparisons with factors that may be used to limit their competitive attributes, environmental impact and spread. In this way specific laboratory, greenhouse and microcosm studies can be related to developmental projections made in more realistic habitats.

Methods: Detailed plant growth and life-table experiments will be conducted on Arundo donax and saltcedar under a range of moisture, nutrient, climate and soil conditions. A combination of highly controlled greenhouse and observational field assessments will be used to gain a mechanistic understanding of the factors that affect plant growth and development and the rate processes that control the speed and magnitude of growth and reproduction. Three-dimensional digitizers will be used to record plant growth in sito and formulated into L-systems models using the methods of Room et al. (1994). Studies will target early stages of plant establishment through both sexual and asexual modes of reproduction, and mature plants in later stages of growth and development, including seed production for saltcedar (giant reed only reproduces vegetatively in North America). The differing developmental phenologies of Tamarix parviflora (early spring bloom) verses that of T. ramosissima (continuous spring/ summer bloom) will be investigated and related to seedbed availability and regional flooding cycles that make new habitats available for both native and invasive species establishment in riparian systems. YST growth and developmental modeling will be conducted in cooperation with researchers at UC Berkeley that already have this work underway (M. Pitcairn personal communication.)

Output/ products: A mechanistic understanding of the physiological and population level growth, development, and reproductive patterns of saltcedar and giant reed. Physiological and structural growth models of YST, saltcedar and giant reed that can be evaluated under a variety of abiotic and biotic growing conditions and used in conjunction with other studies investigating the impact of various measures on their control.

Objective 2b: Investigate the biology and ecology of insect and plant pathogenic natural enemies of the invasive pest plants, develop and assess their target and non-target impacts, and develop methods to use them in integrate weed management programs.

Justification: Although a single biological control agent for saltcedar, Diorhabda elongata, has cleared quarantine and was released into field cage sites in 1999, detailed biological assessments are still needed to characterize different aspects of its biology under natural field conditions. In addition, the biology, host-feeding attributes and other factors need to be investigated for additional natural enemies for all of the target weeds, especially saltcedar. Foreign exploration and host-specificity testing is being conducted using other base-funded resources (see current and pending support), however, assistance though the IFAFS program would highly accelerate efforts to secure regulatory approved release permits for other potential biological control agents. This is an important issue, as multiple biological control agents are often required to exert the necessary levels of stress to control the targeted plant species. Biological studies are also needed on other natural enemies to gain a predictive understanding of their impacts under varying conditions and methods of manipulation and integration with other control factors. Our hypothesis is that additional biological control agents and a mechanistic understanding of their biological attributes will allow us to more effectively develop and use them as key components of an integrated weed management program.

Methods: A combination of laboratory and greenhouse studies on yet to be approved agents will be conducted in the quarantine facilities at Albany, CA, Las Cruces, NM, and Temple, TX. Factors to be investigated included temperature-dependent developmental biology and survival, reproductive capacity, and sampling methodologies including pheromone attraction. For those natural enemies that have already been released from quarantine (e.g. Diorhabda elongata for saltcedar and Chaetorellia succinea for YST), field cage and open field studies will be conducted to determine reproductive rates, mortality rates, behavior and seasonal abundance. The affects of climatic factors will be determined through the use of life-table assessments, key factor analysis and cohort studies.

Output/ products: New regulatory approved biological control agents will be made available for release and use in integrated weed management programs, detailed knowledge of the biology of additional biological control agents, and methods for their control in integrated weed management systems will be developed. Field dispersal studies will provide rates of natural enemy spread and biological control redistribution methods.

Objective 2c: Characterize and evaluate synergistic and antagonistic interactions for various natural enemies, the target pest plant species and abiotic and biotic elements that affect the community level dynamics of riparian ecosystems.
Justification: Invasive plants that dominate riparian and upland habitats often highly alter different attributes of infested ecosystems and the surrounding environment, including biological and physical aspects of the associated community. Their competitive actions also affect and limit the growth and development of other species including beneficial species for agricultural uses and native species that provide improved habitat for desired flora and fauna. Likewise, introduced biological control agents hold the potential to limit invasive species populations but also have the potential to cause non-target effects. These population and community level interactions determine how local food webs are structured and what species (plant and animal) can successfully inhabit an area or can be produced by agricultural means. Native species that currently occupy niches in local habitats can also limit the establishment and spread of natural enemies that may be useful in controlling invasive species. Such may be the case with generalized predators such as ants and spiders that are commonly found in and around riparian zones. Top down suppression of herbivore populations through the actions of oligophagous and polyphagous natural enemies can kill herbivores in areas of introduction and limit biological control programs and need to be assessed in regard to the selection of new natural enemies. It has also been suggested that biological control projects may alter the local fauna and that inadequate pre- and post-release evaluations are conducted. Our hypotheses are that knowledge of the associated habitats, the organisms that inhabit them and their effects on the establishment and effectiveness of introduced biological control agents may further clarify these situations. In addition, this knowledge should help to determine the benefits, risks and other affects that both invasive species and purposefully introduced natural enemies have on agricultural and native ecosystems. Information of this type is essential for Tasks 1 and 3 of this proposal and the field of ecological science and biological control in general.

Methods: Four different aspects of community level interactions will be investigated using a variety of different methods depending upon the circumstances. The first area is plant to plant competitive interactions that will be evaluated at three different scales, including establishment competition for newly available habitats, competitive growth and acquisition of local resources for vegetative expansion and reproduction, and larger scale habitat domination that may occur at landscape levels through alteration of other ecosystem attributes. These will be investigated through a combination of specific experiments and monitoring vegetative patterns over time and space. Second, herbivory such as biological control agent impact on target and non-target plant species will be assessed. The physiological interactions and population level effects of these impacts will be studied using a combination of controlled greenhouse, field cage studies and structural plant growth modeling (see Room et al. 1994). Third, predation and parasitism will be evaluated as to their effects on natural enemy establishment, population expansion and spread. Specifically, predation rates affecting D. elongata and the Tamarix leafhopper Opsuis stactogalus in saltcedar and a predaceous mite affecting C. succinea will be evaluated in YST. For large predators such as the ants affecting D. elongata, DNA markers in the guts of field collected predators will be used to determine predation rates and establishment potential. Lastly, faunistic surveys in both desired vegetation and in habitats that are dominated by invasive species will be assessed both pre and post biological control. Comparative abundance of flora and fauna within and between different habitat types will be compared to aid in the overall evaluation of the benefits and risks of control methods including biologically-based integrated weed management.

Output/ products: The results of this objective will provide information of the population and community level interactions of various native and invasive species, including both beneficial and potentially negative impacts caused by different integrated weed management tactics. Specifically, it will help assess the ecosystem level interactions of biological control agents and provide new data that is being demanded by conservation biologists on the threat of invasive species to threatened and endangered species and other organisms in affected environments. It will also provide significant new data that can be used in Task 1 to more effectively evaluate the data needs of new benefit/ risk assessment methods for biological control regulation.

Objective 2d: Conduct landscape level assessments and predictions of invasive species distribution, assess competition with beneficial species and determine the effects of these pest plant dominated habitats on local and watershed level diversity of associated flora and fauna.

Justification: Invasive species make both local and large-scale impacts on affected agricultural and natural ecosystems. Many times evaluation only occurs at the local level where it is both easy and practical to measure these effects through direct experimentation. Landscape level assessments require additional tools that can provide both wider geographic perspectives and longer time horizons. To accomplish these goals, the Consortium plans to collect both extensive baseline data (see Task 3) over 13 different field sites in six different states and compare them with data collected over the duration of this project and beyond. Although some aerial and ground based methods are currently available for collecting these samples and making these assessments, new remote sensing technologies would make these studies both more accurate and more cost effective. Using two primary activities, we plan to develop new methods that can be used in the implementation aspects of this program and by others interested in the assessment and management of invasive plant species. We hypothesize that new hyperspectral sensor data will provide more accurate assessment technology when evaluated using GIS based models. Similarly, we hypothesize that computer-based plant and natural enemy models implemented spatially through GIS technologies can be used to better assess and predict spatial occurrence, synchronies and spread of key organisms across landscapes. These technologies will be developed and used to aid in the development and evaluation of new biologically-based weed management technologies and in their delivery to local land managers.
Methods: New remote sensing methods will be developed using hyperspectral imaging and standard photographic systems that will be aerially deployed at the research locations in California, Colorado, and/or Texas, and at other specific sites of special interest. A CAS-II imager and a hyperspectral radiometer are already available to the project through G. Anderson of the ARS, Sidney, MT location, and photographic equipment, video imaging systems and imaging aircraft are available through J. Everitt of ARS. Weslaco, TX. Using these remote sensed data, efforts will focus on developing detection, mapping and inventory capabilities for the target weeds and other beneficial plants that are often displaced by invasive species. Specifically, we are interested in detecting the target weeds, willows, cottonwoods and other riparian natives that are displaced by saltcedar/ Arundo, and beneficial forage species that are displaced by YST. These remote sensed data will be used in GIS models both for assessment and prediction of potential areas of pest plant and beneficial species expansion, using various actual and hypothetical management scenarios. Additional data on soil, topography hydrology, geography and other vegetative data that is already available for most of the research sites will be incorporated into these models. These GIS-based models will be constructed and used for the primary purposes of predicting, a) the potential for infested and uninfested areas in the absence of mitigative measures, and b) the potential for existing vegetation to recolonize infested areas following mechanical, herbicidal or biological control of the invasive plant species. Likewise, mechanistic plant growth and natural enemy models will be developed and implemented across spatial grids using new models developed in HERMES (see Carruthers et al. 1992, Legaspi et al. 1997) that will be linked to geographically distributed weather data and GIS delivery systems. Using these techniques new management scenarios can be developed and evaluated prior to actually implementing them in the field. Since large-scale experimentation is extremely expensive, such evaluations will be useful in assessing strategies prior to actual implementation of activities through Task 3.

Output/ products: New remote sensing tools, plant growth and natural enemy models, and GIS/ GPS technologies will be developed, tested through verification and validation procedures and made available to implementation portions of this project and others interested in improved management of invasive weeds.

Objective 2e: Develop re-vegetation strategies, integrate control methods, and test decision models for use in implementing control and re-vegetation under varying environmental conditions.
Justification: As invasive species are controlled, it is essential that desired species can be encouraged to fill the empty niches rather than allowing the reinfestation of the area with regrowth of the same detrimental species or another equally bad or worse invasive species. In agricultural situation, this may be fulfilled with active seeding or through timed management practices that encourage beneficial species over exotic invaders (e.g. correctly timed grazing, burning, or by other cultural practices). In more wild riparian areas, particularly those in arid areas where available moisture is highly limiting, re-vegetation is more difficult and must be carefully evaluated and managed. This is particularly important where altered riparian systems now are almost exclusively populated with exotic invasive species such as saltcedar. In some areas of Arizona, California, New Mexico and Texas, the removal of saltcedar and/ or Arundo donax may cause some threatened and endangered species total loss of local habitat unless effective re-vegetation efforts can be implemented. It is also important that these re-vegetation efforts and other landscape level management actions be properly coordinated and integrated with other management activities. Such integration will require the development of decision aids and other planning and technology transfer methods. It is our hypothesis that re-vegetation and other large-scale management activities can be adequately developed and properly implemented if based on appropriate data and delivered using well developed and tested decision aids. The following three activities are planned to aid in the development, testing and delivery of such tools.

Methods: Specific large-scale study sites (ca 1 km sq.) will be established in habitats of critical concern, particularly in areas such as Texas and New Mexico that closely represents habitats for the southwestern willow flycatcher. Vegetation and wildlife will be characterized and then the target plants will be selectively removed by directed herbicide treatment. Subsequent changes in the diversity and abundance of vegetation, birds and other wildlife will be assessed and compared to adjacent untreated plots. In addition, subplots within the treated and untreated areas will be planted with willows, cottonwoods and other desired species based on existing re-vegetation methods developed by the NRCS Plant Materials Laboratory in Las Lunas, NM. In parallel with these studies, a revegetation decision model will be developed using currently available knowledge, a variety of input variables and control parameters (control method used, soil type, climate, length of growing season, precipitation patterns, flooding frequency and seasonality, presence of desirable and undesirable seed sources, depth to water table, soil salinity, ground water salinity, site management objectives and restoration options), and implemented using precise phasing of factor statements and rank-order scenarios developed by riparian restoration experts. Validation activities will be conducted through tests made by comparing re-vegetation expert recommendations to model output for a series of restoration sites. In all locations, the effects of using several different vegetation control methods such as mechanical, mechanical-chemical, chemical alone, and different cultural and biological control tactics, singly and in combination will be evaluated and integrated into optimal control methods for different local situations. The Bureau of Reclamation will also develop these differing control tactics into a decision support model for local resource and land managers.

Output/ products: New re-vegetation strategies will be developed and existing methods validated for specific sites of most concern to the Consortium goals. A resource manager oriented decision aid will be developed for re-vegetation planning and linked to site specific input data that land managers can use to determine appropriate actions at their local sites. Integrated weed management practices will be linked with biologically-based and other weed management tactics and made available through classical extension delivery channels and through training of land management agency personnel.