US-IALE 2018

AUTHORS: Caroline A. Curtis*, Valerie J. Pasquarella, Bethany A. Bradley – University of Massachusetts

ABSTRACT: The spread of non-native conifers into areas naturally dominated by other vegetation types is a growing problem in South America. This process results in a landscape transformation as the conifers suppress growth and regeneration of native vegetation leading to altered water and nutrient availability and reduced biodiversity. In parts of Chile, pines are problematic due to widespread establishment of plantations and failure to control expanding populations. The large spatial extent and detectable spectral characteristics of non-native pines provide a unique opportunity to apply remotely sensed data to identify patterns of pine occurrences across the landscape. Previous research has identified the current and historical extent of non-native pines to quantify land cover changes. However, most studies have used a limited number of images to characterize change and covered a relatively small spatial area. This research uses a dense time series of Landsat images which allows us to capture the gradual invasion and dispersal process over more of the affected area. We used all available Landsat images for two scenes (Path 231/Row 092 and Path 232/Row 092) in southern Chile. For each scene, we created training data based on historical aerial photos from Google Earth. We then used time series models to quantify when land cover changes occurred and the Random Forest algorithm to classify images and create biennial maps showing changes in land cover. This method allowed us to use all of the high-quality pixels from each image, greatly improving our understanding of the spatial and temporal patterns of land cover change. In addition to creating risk maps for conifer invasion in Chile, this research also has implications for other ecosystems as it highlights the importance of using dense time series to monitor conifer spread.

AUTHORS: Gabriela Nunez-Mir*, Songlin Fei – Purdue University

ABSTRACT: Empirical and theoretical studies have shown that the association between native diversity and invasion (e.g., usually exotic richness) is inconsistent, varying across scales and ecosystems of study. To explain the different patterns, ecologists have proposed a number of underlying processes, focusing mostly on the role of local or regional native species richness. Yet, there are other aspects of native diversity involved in influencing a community’s invasibility that may be contributing to the inconsistencies observed. Although abiotic environmental heterogeneity has been extensively explored as a driver of these patterns, little is known about the role of biotic heterogeneity. Here, we seek to understand how beta diversity, the heterogeneity of species composition among assemblages, influences the observed relationship between native and exotic species richness. In addition, we sought to understand how this influence changes across scales. To do this, we obtained native and exotic richness data for the entire continental United States from the USGS nonindigenous aquatic species and NatureServe databases. We took advantage of the nested properties of these databases (HUC levels) to incorporate the influence of scale in these relationships. Because beta diversity is likely to be a good indicator for spatial heterogeneity, we hypothesized that high beta diversity enhances coexistence and supersedes processes that decrease invasibility. Therefore, we expected that areas with higher levels of beta diversity would display more positive relationships. Our results seem to indicate a scale-dependent effect of beta diversity on invasibility. Contrary to our hypothesis, we found beta diversity to have, not a negative effect, but a positive effect on invasibility, and only at scales with the largest spatial extent. Ultimately, our study contributes to our understanding of these invasion patterns and their relationship with scale, as well as to our ability to evaluate the vulnerability of communities to invasion.

AUTHORS: Kathryn C. Baer *, Andrew N. Gray – US Forest Service Pacific Northwest Research Station

ABSTRACT: A central question in landscape ecology concerns the environmental determinants of species distributions and the scale at which these determinants act. This question is of particular importance when attempting to predict the distribution of invasive species. Biogeographical theory upon which distribution modeling techniques are built predicts that abiotic constraints are of primary importance at the landscape scale, while biotic resistance or facilitation are likely to primarily impact a species’ local distribution. However, tests of this theory have yielded equivocal results. In this study, we utilized observational data collected on over 9,000 plots from 2004-2011 in Washington, Oregon, and California to identify and evaluate the relative importance of correlates of the presence of Bromus tectorum, Cirsium vulgare, and Rubus armeniacus across spatial scales. Contrary to predictions that biotic constraints should not affect distribution at this level, the landscape-scale presence of all three invasive species was strongly correlated with aspects of both the biotic and abiotic environment. Proximity to the nearest road was more strongly predictive of C. vulgare and R. armenicus presence at the landscape than the local scale, indicating dispersal limitation of these species’ geographic distributions. Several abiotic conditions such as annual precipitation showed a stronger relationship with local- than landscape-scale patterns of invasive species presence. Aspects of the local biotic environment including forest vegetation cover and grazing history were also more important at the local level, providing a measure of support for the prediction of biogeographical theory that interactions such as biotic resistance and herbivory are likely to be more important in determining local than landscape-level distributions. Our results suggest that the incorporation of biotic variables and dispersal limitation into the construction of distribution models, particularly for invasive species, is likely substantially improve the accuracy of their predictions.

AUTHORS: Insu Jo*, Purdue University; Kevin Potter, North Carolina State University; Grant Domke, USDA Forest Service; Songlin Fei, Purdue University

ABSTRACT: Plant-fungal interactions play an important role in forest community structure and functions; however, the dominant tree-fungal interactions that affect species invasions in forest understories remain largely unexplored. Using Forest Inventory Analysis program data from the US Forest Service, we examined how dominant tree mycorrhizal type affects forest soil properties, and in turn, how the altered forest soil properties and forest structure affect understory plant invasions in eastern US forests. We found that understory invaders were more abundant in arbuscular mycorrhizal (AM) dominant forests than ectomycorrhizal (ECM) dominant forests, whereas native species cover and richness had no strong associations with AM tree dominance. Among the soil attributes that were significantly influenced by AM tree dominance, forest floor thickness and soil carbon:nitrogen ratio were significantly associated with invader richness and cover. Our results suggest that forest structure and dominant mycorrhizal type are closely linked with understory plant invasions. The increased invader abundance in AM dominant forests can further facilitate nutrient cycling, such as nitrogen cycling, altering ecosystem structure and functions in forest ecosystems.

AUTHORS: Jake Kvistad*, University of Toledo; Jonathan Bossenbroek, University of Toledo; Lindsay Chadderton, The Nature Conservancy

ABSTRACT: Ballast water is a major vector for aquatic invasive species (AIS) introduction and spread on the Laurentian Great Lakes. Understanding inter-lake patterns of domestic ballast water exchanges on the Great Lakes will aid in designing effective ballast water management plans that can either slow or prohibit secondary spread of potentially harmful AIS. Graph theoretic analysis techniques were used to identify central ports in a network of 151 Great Lakes ports that should be considered for increased management efforts. Twenty-seven ports which scored high for several network centrality metrics were identified for hypothetical management focus. Aquatic invasive species secondary spread patterns were simulated for 1000 trials of both 3 and 10 time-steps using a stochastic model of basin-wide ballast water exchanges between ports under a continuum of management scenarios ranging from “no management” to “100% management” at 10% incremental increases of potential management effort. Infestation probabilities resulting from 10 time-steps of simulated spread in both 100% managed and unmanaged scenarios differed significantly (p < 0.05), and average basin-wide infestation probabilities reduced by 75.6%. Average basin-wide infestation probabilities after 10 time-steps decreased exponentially as management effort increased, with the greatest reductions in infestation probabilities occurring between 70 and 100% management effort. Average basin-wide infestation probabilities after 3 time-steps decreased approximately linearly. Differences in average infestation probabilities after 3 time-steps 10 time-steps across management scenarios suggest that management loses effectiveness the longer an organism is allowed to spread. Some ports within the Great Lakes basin, specifically in northern Lakes Michigan and Huron, did not show any appreciable reduction in infestation probabilities, suggesting additional scrutiny is required for AIS management in these areas. These simulations show that targeted management at central ports within the Great Lakes basin have the potential to dramatically reduce and slow AIS secondary spread through ballast water exchange.

AUTHORS: Whalen W. Dillon*, University of Florida; Brian F. Allan, University of Illinois; Michael C. Dietze, Boston University; S. Luke Flory, University of Florida

ABSTRACT: Ecological theory predicts that climate change will alter vector-borne disease risk through a variety of direct and indirect pathways, but the relative importance of various pathways is poorly understood for most systems. We are exploring several factors hypothesized to affect the risk of exposure to tick-borne diseases (TBDs) on Department of Defense (DoD) installations along a gradient of climatic conditions across the southeastern United States. Specifically, we are examining the effects of fire regimes and plant invasions on abundances of ticks and wildlife hosts, focusing on non-native cogongrass (Imperata cylindrica) and the lone star tick (Amblyomma americanaum). Our overarching hypothesis is that fire and plant invasions are dominant factors directly and indirectly affecting exposure risk to TBDs on DoD installations in this region. In the first year of this multi-year study we visited six DoD installations, collecting data on forest structure, the plant community, and abundances of ticks and wildlife hosts at 29 plots primarily in pine-dominated landscapes varying from 0 to 11 years since the last prescribed fire. Cogongrass management efficacy on DoD installations limited the number of untreated invaded areas to eight plots, with six of these burned less than three years prior. Tick-host abundance was greater at 0-1 years than 2+ years, following fire. Tick abundance was greater in uninvaded plots, increasing with time since fire, litter cover, and over-story canopy cover. Climate effects based on the latitudinal gradient of installations weren’t detectable in our preliminary data set. Our initial results provide evidence suggesting that fire and forest vegetation and structure have significant effects on TBD risk as we predicted. Frequent prescribed fires are a typical management goal to maintain pine ecosystems, and we show the additional benefit of lowering TBD risk by reducing canopy and litter cover.

AUTHORS: Samantha Schwab*, Rutgers University; Chris Stone, University of Illinois at Urbana-Champaign; Dina Fonseca, Center for Vector Biology; Nina Fefferman, University of Tennessee Knoxville

ABSTRACT: Recent epidemics of mosquito-borne dengue and Zika viruses demonstrate the urgent need for effective measures to control these diseases. The best method currently available to prevent or reduce the size of outbreaks is to reduce the abundance of their mosquito vectors, but there is little consensus on which mechanisms of control are most effective, or when and where they should be implemented. Although the optimal methods are likely context-dependent, broadly applicable strategies for mosquito control, such as how to distribute limited resources across a landscape in times of high epidemic risk, can mitigate (re)-emerging outbreaks. We used mathematical simulations to examine how the spatial distribution of larval mosquito control affects the size of disease outbreaks, and how mosquito metapopulation dynamics and demography (ecological context) might impact the efficacy of different spatial distributions of control. We found that the birth rate and mechanism of density-dependent regulation of mosquito populations affected the average outbreak size across all control distributions, with a higher birth rate leading to smaller outbreaks in some circumstances. Ecological context also determined whether spatial control distributions favoring the interior or the edges of the landscape most effectively reduced human infections. Thus, control efforts that consider local mosquito population dynamics to determine how to distribute limited resources across space may reduce the size of mosquito-borne disease outbreaks more effectively than uninformed efforts.

AUTHORS: Matt Unitis*, Jodi Brandt – Boise State University

ABSTRACT: Great Lakes coastal wetlands have an ecological importance disproportionate to their size, yet are extremely vulnerable to invasion by introduced plants. Plant invasions can be effectively controlled in early stages of invasion, but early detection of invasive wetland plants using traditional ground survey methods is time-prohibitive, and remote detection is challenging because of the heterogeneity of plant composition during early stages of invasion. New satellite data sources and analysis techniques offer promise for wetland and invasive plant remote sensing, and in this study we used Sentinel-1 C-band SAR from three different dates with RapidEye multispectral optical imagery to detect phenological and spectral differences between wetland plant communities in the Great Lakes. Of particular interest was detecting hybrid cattail (Typha x. glauca), which imperils the high quality coastal wetlands of the St. Mary’s River. We compared wetland classifications using Randomforest and Maximum Likelihood algorithms, and subsequently determined the best Randomforest model using various data combinations of optical and radar inputs. Both techniques produced classifications of similar accuracy (Random Forest achieved a 94.0% average overall accuracy and Maximum Likehood achieved 90.1% overall accuracy). The best-performing classification was a Randomforest model with multi-temporal SAR and optical bands as input variables, achieving a 97.4% producer’s accuracy for hybrid cattail. Randomforest variable importance demonstrated that the Near Infrared band of RapidEye and the August 22nd SAR were the two most important variables to discriminate between wetland plant communities. Our results indicate that by combining moderate resolution SAR and optical datasets, we can map detailed wetland communities at a scale and resolution to guide early detection and management of wetland plant invasions. Furthermore, our methods can be used for repeat monitoring of hybrid cattail, and are likely adaptable for remote monitoring in other wetland environments due to our exclusive use of spaceborne platforms with global coverage.