A modeling workflow that balances automation and human intervention to inform invasive plant management decisions at multiple spatial scales.

Predictions of habitat suitability for invasive plant species can guide risk assessments at regional and national scales and inform early detection and rapid-response strategies at local scales. We present a general approach to invasive species modeling and mapping that meets objectives at multiple scales. Our methodology is designed to balance trade-offs between developing highly customized models for few species versus fitting non-specific and generic models for numerous species. We developed a national library of environmental variables known to physiologically limit plant distributions and relied on human input based on natural history knowledge to further narrow the variable set for each species before developing habitat suitability models. To ensure efficiency, we used largely automated modeling approaches and human input only at key junctures. We explore and present uncertainty by using two alternative sources of background samples, including five statistical algorithms, and constructing model ensembles. We demonstrate the use and efficiency of the Software for Assisted Habitat Modeling [SAHM 2.1.2], a package in VisTrails, which performs the majority of the modeling analyses. Our workflow includes solicitation of expert feedback on model outputs such as spatial prediction results and variable response curves, and iterative improvement based on new data availability and directed field validation of initial model results. We highlight the utility of the models for decision-making at regional and local scales with case studies of two plant species that invade natural areas: fountain grass (Pennisetum setaceum) and goutweed (Aegopodium podagraria). By balancing model automation with human intervention, we can efficiently provide land managers with mapped predicted distributions for multiple invasive species to inform decisions across spatial scales.

Integrating Remote Sensing with Species Distribution Models; Mapping Tamarisk Invasions Using the Software for Assisted Habitat Modeling (SAHM).

Early detection of invasive plant species is vital for the management of natural resources and protection of ecosystem processes. The use of satellite remote sensing for mapping the distribution of invasive plants is becoming more common, however conventional imaging software and classification methods have been shown to be unreliable. In this study, we test and evaluate the use of five species distribution model techniques fit with satellite remote sensing data to map invasive tamarisk (Tamarix spp.) along the Arkansas River in Southeastern Colorado. The models tested included boosted regression trees (BRT), Random Forest (RF), multivariate adaptive regression splines (MARS), generalized linear model (GLM), and Maxent. These analyses were conducted using a newly developed software package called the Software for Assisted Habitat Modeling (SAHM). All models were trained with 499 presence points, 10,000 pseudo-absence points, and predictor variables acquired from the Landsat 5 Thematic Mapper (TM) sensor over an eight-month period to distinguish tamarisk from native riparian vegetation using detection of phenological differences. From the Landsat scenes, we used individual bands and calculated Normalized Difference Vegetation Index (NDVI), Soil-Adjusted Vegetation Index (SAVI), and tasseled capped transformations. All five models identified current tamarisk distribution on the landscape successfully based on threshold independent and threshold dependent evaluation metrics with independent location data. To account for model specific differences, we produced an ensemble of all five models with map output highlighting areas of agreement and areas of uncertainty. Our results demonstrate the usefulness of species distribution models in analyzing remotely sensed data and the utility of ensemble mapping, and showcase the capability of SAHM in pre-processing and executing multiple complex models.

Evidence of niche shift and global invasion potential of the Tawny Crazy ant, Nylanderia fulva.

Analysis of an invasive species' niche shift between native and introduced ranges, along with potential distribution maps, can provide valuable information about its invasive potential. The tawny crazy ant, Nylanderia fulva, is a rapidly emerging and economically important invasive species in the southern United States. It is originally from east-central South America and has also invaded Colombia and the Caribbean Islands. Our objectives were to generate a global potential distribution map for N. fulva, identify important climatic drivers associated with its current distribution, and test whether N. fulva's realized climatic niche has shifted across its invasive range. We used MaxEnt niche model to map the potential distribution of N. fulva using its native and invaded range occurrences and climatic variables. We used principal component analysis methods for investigating potential shifts in the realized climatic niche of N. fulva during invasion. We found strong evidence for a shift in the realized climatic niche of N. fulva across its invasive range. Our models predicted potentially suitable habitat for N. fulva in the United States and other parts of the world. Our analyses suggest that the majority of observed occurrences of N. fulva in the United States represent stabilizing populations. Mean diurnal range in temperature, degree days at ≥10°C, and precipitation of driest quarter were the most important variables associated with N. fulva distribution. The climatic niche expansion demonstrated in our study may suggest significant plasticity in the ability of N. fulva to survive in areas with diverse temperature ranges shown by its tolerance for environmental conditions in the southern United States, Caribbean Islands, and Colombia. The risk maps produced in this study can be useful in preventing N. fulva's future spread, and in managing and monitoring currently infested areas.

Using high-resolution future climate scenarios to forecast Bromus tectorum invasion in Rocky Mountain National Park.

National Parks are hallmarks of ecosystem preservation in the United States. The introduction of alien invasive plant species threatens protection of these areas. Bromus tectorum L. (commonly called downy brome or cheatgrass), which is found in Rocky Mountain National Park (hereafter, the Park), Colorado, USA, has been implicated in early spring competition with native grasses, decreased soil nitrogen, altered nutrient and hydrologic regimes, and increased fire intensity. We estimated the potential distribution of B. tectorum in the Park based on occurrence records (n = 211), current and future climate, and distance to roads and trails. An ensemble of six future climate scenarios indicated the habitable area of B. tectorum may increase from approximately 5.5% currently to 20.4% of the Park by the year 2050. Using ordination methods we evaluated the climatic space occupied by B. tectorum in the Park and how this space may shift given future climate change. Modeling climate change at a small extent (1,076 km2) and at a fine spatial resolution (90 m) is a novel approach in species distribution modeling, and may provide inference for microclimates not captured in coarse-scale models. Maps from our models serve as high-resolution hypotheses that can be improved over time by land managers to set priorities for surveys and removal of invasive species such as B. tectorum.

No universal scale-dependent impacts of invasive species on native plant species richness.

A growing number of studies seeking generalizations about the impact of plant invasions compare heavily invaded sites to uninvaded sites. But does this approach warrant any generalizations? Using two large datasets from forests, grasslands and desert ecosystems across the conterminous United States, we show that (i) a continuum of invasion impacts exists in many biomes and (ii) many possible species-area relationships may emerge reflecting a wide range of patterns of co-occurrence of native and alien plant species. Our results contradict a smaller recent study by Powell et al. 2013 (Science 339, 316-318. (doi:10.1126/science.1226817)), who compared heavily invaded and uninvaded sites in three biomes and concluded that plant communities invaded by non-native plant species generally have lower local richness (intercepts of log species richness-log area regression lines) but steeper species accumulation with increasing area (slopes of the regression lines) than do uninvaded communities. We conclude that the impacts of plant invasions on plant species richness are not universal.

The hyper-envelope modeling interface (HEMI): a novel approach illustrated through predicting tamarisk (Tamarix spp.) habitat in the Western USA.

Habitat suitability maps are commonly created by modeling a species' environmental niche from occurrences and environmental characteristics. Here, we introduce the hyper-envelope modeling interface (HEMI), providing a new method for creating habitat suitability models using Bezier surfaces to model a species niche in environmental space. HEMI allows modeled surfaces to be visualized and edited in environmental space based on expert knowledge and does not require absence points for model development. The modeled surfaces require relatively few parameters compared to similar modeling approaches and may produce models that better match ecological niche theory. As a case study, we modeled the invasive species tamarisk (Tamarix spp.) in the western USA. We compare results from HEMI with those from existing similar modeling approaches (including BioClim, BioMapper, and Maxent). We used synthetic surfaces to create visualizations of the various models in environmental space and used modified area under the curve (AUC) statistic and akaike information criterion (AIC) as measures of model performance. We show that HEMI produced slightly better AUC values, except for Maxent and better AIC values overall. HEMI created a model with only ten parameters while Maxent produced a model with over 100 and BioClim used only eight. Additionally, HEMI allowed visualization and editing of the model in environmental space to develop alternative potential habitat scenarios. The use of Bezier surfaces can provide simple models that match our expectations of biological niche models and, at least in some cases, out-perform more complex approaches.

Regional data refine local predictions: modeling the distribution of plant species abundance on a portion of the central plains.

Species distribution models are frequently used to predict species occurrences in novel conditions, yet few studies have examined the consequences of extrapolating locally collected data to regional landscapes. Similarly, the process of using regional data to inform local prediction for species distribution models has not been adequately evaluated. Using boosted regression trees, we examined errors associated with extrapolating models developed with locally collected abundance data to regional-scale spatial extents and associated with using regional data for predictions at a local extent for a native and non-native plant species across the northeastern central plains of Colorado. Our objectives were to compare model results and accuracy between those developed locally and extrapolated regionally, those developed regionally and extrapolated locally, and to evaluate extending species distribution modeling from predicting the probability of presence to predicting abundance. We developed models to predict the spatial distribution of plant species abundance using topographic, remotely sensed, land cover and soil taxonomic predictor variables. We compared model predicted mean and range abundance values to observed values between local and regional. We also evaluated model prediction performance based on Pearson's correlation coefficient. We show that: (1) extrapolating local models to regional extents may restrict predictions, (2) regional data can help refine and improve local predictions, and (3) boosted regression trees can be useful to model and predict plant species abundance. Regional sampling designed in concert with large sampling frameworks such as the National Ecological Observatory Network may improve our ability to monitor changes in local species abundance.

Federated or cached searches: Providing expected performance from multiple invasive species databases.

Invasive species are a universal global problem, but the information to identify them, manage them, and prevent invasions is stored around the globe in a variety of formats. The Global Invasive Species Information Network is a consortium of organizations working toward providing seamless access to these disparate databases via the Internet. A distributed network of databases can be created using the Internet and a standard web service protocol. There are two options to provide this integration. First, federated searches are being proposed to allow users to search "deep" web documents such as databases for invasive species. A second method is to create a cache of data from the databases for searching. We compare these two methods, and show that federated searches will not provide the performance and flexibility required from users and a central cache of the datum are required to improve performance.

Ensemble habitat mapping of invasive plant species.

Ensemble species distribution models combine the strengths of several species environmental matching models, while minimizing the weakness of any one model. Ensemble models may be particularly useful in risk analysis of recently arrived, harmful invasive species because species may not yet have spread to all suitable habitats, leaving species-environment relationships difficult to determine. We tested five individual models (logistic regression, boosted regression trees, random forest, multivariate adaptive regression splines (MARS), and maximum entropy model or Maxent) and ensemble modeling for selected nonnative plant species in Yellowstone and Grand Teton National Parks, Wyoming; Sequoia and Kings Canyon National Parks, California, and areas of interior Alaska. The models are based on field data provided by the park staffs, combined with topographic, climatic, and vegetation predictors derived from satellite data. For the four invasive plant species tested, ensemble models were the only models that ranked in the top three models for both field validation and test data. Ensemble models may be more robust than individual species-environment matching models for risk analysis.

Global potential net primary production predicted from vegetation class, precipitation, and temperature.

Net primary production (NPP), the difference between CO2 fixed by photosynthesis and CO2 lost to autotrophic respiration, is one of the most important components of the carbon cycle. Our goal was to develop a simple regression model to estimate global NPP using climate and land cover data. Approximately 5600 global data points with observed mean annual NPP, land cover class, precipitation, and temperature were compiled. Precipitation was better correlated with NPP than temperature, and it explained much more of the variability in mean annual NPP for grass- or shrub-dominated systems (r2 = 0.68) than for tree-dominated systems (r2 = 0.39). For a given precipitation level, tree-dominated systems had significantly higher NPP (approximately 100-150 g C m(-2) yr(-1)) than non-tree-dominated systems. Consequently, previous empirical models developed to predict NPP based on precipitation and temperature (e.g., the Miami model) tended to overestimate NPP for non-tree-dominated systems. Our new model developed at the National Center for Ecological Analysis and Synthesis (the NCEAS model) predicts NPP for tree-dominated systems based on precipitation and temperature; but for non-tree-dominated systems NPP is solely a function of precipitation because including a temperature function increased model error for these systems. Lower NPP in non-tree-dominated systems is likely related to decreased water and nutrient use efficiency and higher nutrient loss rates from more frequent fire disturbances. Late 20th century aboveground and total NPP for global potential native vegetation using the NCEAS model are estimated to be approximately 28 Pg and approximately 46 Pg C/yr, respectively. The NCEAS model estimated an approximately 13% increase in global total NPP for potential vegetation from 1901 to 2000 based on changing precipitation and temperature patterns.

The myth of plant species saturation.

Plant species assemblages, communities or regional floras might be termed 'saturated' when additional immigrant species are unsuccessful at establishing due to competitive exclusion or other inter-specific interactions, or when the immigration of species is off-set by extirpation of species. This is clearly not the case for state, regional or national floras in the USA where colonization (i.e. invasion by exotic species) exceeds extirpation by roughly a 24 to 1 margin. We report an alarming temporal trend in plant invasions in the Pacific Northwest over the past 100 years whereby counties highest in native species richness appear increasingly invaded over time. Despite the possibility of some increased awareness and reporting of native and exotic plant species in recent decades, historical records show a significant, consistent long-term increase in exotic species (number and frequency) at county, state and regional scales in the Pacific Northwest. Here, as in other regions of the country, colonization rates by exotic species are high and extirpation rates are negligible. The rates of species accumulation in space in multi-scale vegetation plots may provide some clues to the mechanisms of the invasion process from local to national scales.

Rapid assessment of postfire plant invasions in coniferous forests of the western United States.

Fire is a natural part of most forest ecosystems in the western United States, but its effects on nonnative plant invasion have only recently been studied. Also, forest managers are engaging in fuel reduction projects to lessen fire severity, often without considering potential negative ecological consequences such as nonnative plant species introductions. Increased availability of light, nutrients, and bare ground have all been associated with high-severity fires and fuel treatments and are known to aid in the establishment of nonnative plant species. We use vegetation and environmental data collected after wildfires at seven sites in coniferous forests in the western United States to study responses of nonnative plants to wildfire. We compared burned vs. unburned plots and plots treated with mechanical thinning and/or prescribed burning vs. untreated plots for nonnative plant species richness and cover and used correlation analyses to infer the effect of abiotic site conditions on invasibility. Wildfire was responsible for significant increases in nonnative species richness and cover, and a significant decrease in native cover. Mechanical thinning and prescribed fire fuel treatments were associated with significant changes in plant species composition at some sites. Treatment effects across sites were minimal and inconclusive due to significant site and site x treatment interaction effects caused by variation between sites including differences in treatment and fire severities and initial conditions (e.g., nonnative species sources). We used canonical correspondence analysis (CCA) to determine what combinations of environmental variables best explained patterns of nonnative plant species richness and cover. Variables related to fire severity, soil nutrients, and elevation explained most of the variation in species composition. Nonnative species were generally associated with sites with higher fire severity, elevation, percentage of bare ground, and lower soil nutrient levels and lower canopy cover. Early assessments of postfire stand conditions can guide rapid responses to nonnative plant invasions.

The art and science of weed mapping.

Land managers need cost-effective and informative tools for non-native plant species management. Many local, state, and federal agencies adopted mapping systems designed to collect comparable data for the early detection and monitoring of non-native species. We compared mapping information to statistically rigorous, plot-based methods to better understand the benefits and compatibility of the two techniques. Mapping non-native species locations provided a species list, associated species distributions, and infested area for subjectively selected survey sites. The value of this information may be compromised by crude estimates of cover and incomplete or biased estimations of species distributions. Incorporating plot-based assessments guided by a stratified-random sample design provided a less biased description of non-native species distributions and increased the comparability of data over time and across regions for the inventory, monitoring, and management of non-native and native plant species.

Risk analysis for biological hazards: what we need to know about invasive species.

Risk analysis for biological invasions is similar to other types of natural and human hazards. For example, risk analysis for chemical spills requires the evaluation of basic information on where a spill occurs; exposure level and toxicity of the chemical agent; knowledge of the physical processes involved in its rate and direction of spread; and potential impacts to the environment, economy, and human health relative to containment costs. Unlike typical chemical spills, biological invasions can have long lag times from introduction and establishment to successful invasion, they reproduce, and they can spread rapidly by physical and biological processes. We use a risk analysis framework to suggest a general strategy for risk analysis for invasive species and invaded habitats. It requires: (1) problem formation (scoping the problem, defining assessment endpoints); (2) analysis (information on species traits, matching species traits to suitable habitats, estimating exposure, surveys of current distribution and abundance); (3) risk characterization (understanding of data completeness, estimates of the "potential" distribution and abundance; estimates of the potential rate of spread; and probable risks, impacts, and costs); and (4) risk management (containment potential, costs, and opportunity costs; legal mandates and social considerations and information science and technology needs).

Spatial heterogeneity influences native and nonnative plant species richness.

Spatial heterogeneity may have differential effects on the distribution of native and nonnative plant species richness. We examined the effects of spatial heterogeneity on native and nonnative plant species richness distributions in the central part of Rocky Mountain National Park, Colorado, USA. Spatial heterogeneity around vegetation plots was characterized using landscape metrics, environmental/topographic variables (slope, aspect, elevation, and distance from stream or river), and soil variables (nitrogen, clay, and sand). The landscape metrics represented five components of landscape heterogeneity and were measured at four spatial extents (within varying radii of 120, 240, 480, and 960 m) using the FRAGSTATS landscape pattern analysis program. Akaike's Information Criterion adjusted for small sample size (AICc) was used to select the best models from a set of multiple linear regression models developed for native and nonnative plant species richness at four spatial extents and three levels of ecological hierarchy (i.e., landscape, land cover, and community). Both native and nonnative plant species richness were positively correlated with edge density, Simpson's diversity index and interspersion/juxtaposition index, and were negatively correlated with mean patch size. The amount of variation explained at four spatial extents and three hierarchical levels ranged from 30% to 70%. At the landscape level, the best models explained 43% of the variation in native plant species richness and 70% of the variation in nonnative plant species richness (240-m extent). In general, the amount of variation explained was always higher for nonnative plant species richness, and the inclusion of landscape metrics always significantly improved the models. The best models explained 66% of the variation in nonnative plant species richness for both the conifer land cover type and lodgepole pine community. The relative influence of the components of spatial heterogeneity differed for native and nonnative plant species richness and varied with the spatial extent of analysis and levels of ecological hierarchy. The study offers an approach to quantify spatial heterogeneity to improve models of plant biodiversity. The results demonstrate that ecologists must recognize the importance of spatial heterogeneity in managing native and nonnative plant species.

Assessing vulnerability to invasion by nonnative plant species at multiple spatial scales.

Basic information on where nonnative plant species have successfully invaded is lacking. We assessed the vulnerability of 22 vegetation types (25 sets of four plots in nine study areas) to nonnative plant invasions in the north-central United States. In general, habitats with high native species richness were more heavily invaded than species-poor habitats, low-elevation areas were more invaded than high-elevation areas, and riparian zones were more invaded than nearby upland sites. For the 100 1000-m2 plots (across all vegetation types), 50% of the variation in nonnative species richness was explained by longitude, latitude, native plant species richness, soil total percentage nitrogen, and mean maximum July temperature (n = 100 plots; P < 0.001). At the vegetation-type scale (n = 25 sets of four 1000-m2 plots/type), 64% of the variation in nonnative species richness was explained by native plant species richness, elevation, and October to June precipitation (P < 0.001). The foliar cover of nonnative species (log) was strongly positively correlated with the nonnative species richness at the plot scale (r = 0.77, P < 0.001) and vegetation-type scale (r = 0.83, P < 0.001). We concluded that, at the vegetation-type and regional scales in the north-central United States, (1) vegetation types rich in native species are often highly vulnerable to invasion by nonnative plant species; (2) where several nonnative species become established, nonnative species cover can substantially increase; (3) the attributes that maintain high native plant species richness (high light, water, nitrogen, and temperatures) also help maintain nonnative plant species richness; and (4) more care must be taken to preserve native species diversity in highly vulnerable habitats.

Effects of river level fluctuation on plant species richness, diversity, and distribution in a floodplain forest in Central Amazonia.

River levels in Central Amazonia fluctuate up to 14 m annually, with the flooding period ranging from 50 to 270 days between the rising and falling phases. Vast areas of forest along the rivers contain plant species that are well adapted to annual flooding. We studied the effect of flooding level on tree species richness, diversity, density, and composition in lake, river, and stream habitats in Jaú National Park, Brazil. 3051 trees >10 cm diameter (at 1.3 m diameter at breast height, dbh) were measured and identified in 25 10 m × 40 m randomly selected plots in each habitat. Ordination methods and analysis of variance results showed that forested areas near lakes had significantly lower species richness of trees than riverine and streamside habitats. Plot species richness and diversity were strongly negatively correlated with the water level and duration of flooding. The drier (stream) habitat had more total species (54 species of trees) and more unique species of trees (6 tree species) than the riverine (52 tree species; 3 unique species) and lake (33 tree species; 3 unique species) habitats. Species composition overlap among habitats was surprisingly high (42.6-60.6% overlap), almost one-third of the species were found in all three habitat types, and few species were unique to each habitat. We conclude that: (1) duration of flooding has a strong impact on species richness, diversity and plant distribution patterns; (2) most species are adapted to a wide range of habitats and flood durations; and (3) while flood duration may decrease local diversity, it also creates and maintains high landscape-scale diversity by increasing landscape heterogeneity.