Search USGS⌕ Search

SEARCH · Search USGS

Results for “Applied Vegetation Science”

Search indexed USGS publications on groundwater, aquifers, geologic maps, mineral resources and earthquakes. Explore source records by subject and place.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 37 records · Page 2Linked to original sources
USGS2002

Application of decision-tree techniques to forest group and basal area mapping using satellite imagery and forest inventory data

Accurate, current, and cost-effective fire fuel data are required by management and fire science communities for use in reducing wildland fire hazards over large areas. In this paper we present results of applying decision-tree techniques to mapping vegetation parameters (such as vegetation types and canopy structure classification) required for fire fuel characterization. Specifically, we present preliminary results of mapping forest types and average basal area by different forest types at 30-meter resolution. Input data into the decision tree model included Landsat-7 ETM+ spring, summer and fall greenness, brightness and wetness of the tasseled cap transformation, topographic data layers such as slope and elevation, and forest variables measured on inventory plots in the Mid-Atlantic region. Using decision-tree models, eight forest types were successfully identified in training cases and mapped for the entire mapping area. Forest basal area per unit area (conifer and deciduous) was estimated as well using regression tree models. Cross-validation conducted for both forest types and basal area showed that discrete forest type estimation error was 35% and continuous basal area relative errors were between 58 and 72%. Accuracy was higher in homogeneous forested lands and lower in areas with fragmented forest cover. The study demonstrated that decision tree and regression tree methods are efficient for large-area vegetation mapping if sufficient large-amount of reference data are available.

Conference Paper↗
USGS2024

The spatially adaptable filter for error reduction (SAFER) process: Remote sensing-based LANDFIRE disturbance mapping updates

LANDFIRE (LF) has been producing periodic spatially explicit vegetation change maps (i.e., LF disturbance products) across the entire United States since 1999 at a 30 m spatial resolution. These disturbance products include data products produced by various fire programs, field-mapped vegetation and fuel treatment activity (i.e., events) submissions from various agencies, and disturbances detected by the U.S. Geological Survey Earth Resources Observation and Science (EROS)-based Remote Sensing of Landscape Change (RSLC) process. The RSLC process applies a bi-temporal change detection algorithm to Landsat satellite-based seasonal composites to generate the interim disturbances that are subsequently reviewed by analysts to reduce omission and commission errors before ingestion them into LF’s disturbance products. The latency of the disturbance product is contingent on timely data availability and analyst review. This work describes the development and integration of the Spatially Adaptable Filter for Error Reduction (SAFER) process and other error and latency reduction improvements to the RSLC process. SAFER is a random forest-based supervised classifier and uses predictor variables that are derived from multiple years of pre- and post-disturbance Landsat band observations. Predictor variables include reflectance, indices, and spatial contextual information. Spatial contextual information that is unique to each contiguous disturbance region is parameterized as Z scores using differential observations of the disturbed regions with its undisturbed neighbors. The SAFER process was prototyped for inclusion in the RSLC process over five regions within the conterminous United States (CONUS) and regional model performance, evaluated using 2016 data. Results show that the inclusion of the SAFER process increased the accuracies of the interim disturbance detections and thus has potential to reduce the time needed for analyst review. LF does not track the time taken by each analyst for each tile, and hence, the relative effort saved was parameterized as the percentage of 30 m pixels that are correctly classified in the SAFER outputs to the total number of pixels that are incorrectly classified in the interim disturbance and are presented. The SAFER prototype outputs showed that the relative analysts’ effort saved could be over 95%. The regional model performance evaluation showed that SAFER’s performance depended on the nature of disturbances and availability of cloud-free images relative to the time of disturbances. The accuracy estimates for CONUS were inferred by comparing the 2017 SAFER outputs to the 2017 analyst-reviewed data. As expected, the SAFER outputs had higher accuracies compared to the interim disturbances, and CONUS-wide relative effort saved was over 92%. The regional variation in the accuracies and effort saved are discussed in relation to the vegetation and disturbance type in each region. SAFER is now operationally integrated into the RSLC process, and LANDFIRE is well poised for annual updates, contingent on the availability of data.

Fire↗
USGS2021

Identifying policy-relevant indicators for assessing landscape vegetation patterns to inform planning and management on multiple use public lands

Understanding the structure and composition of landscapes can empower agencies to effectively manage public lands for multiple uses while sustaining land health. Many landscape metrics exist, but they are not often used in public land decision-making. Our objectives were to (1) develop and (2) apply a process for identifying a core set of indicators that public land managers can use to understand landscape-level resource patterns on and around public lands. We first developed a process for identifying indicators that are grounded in policy, feasible to quantify using existing data and resources, and useful for managers. We surveyed landscape monitoring efforts by other agencies, gathered science and agency input on monitoring goals, and quantified the prevalence of potential indicators in agency land health standards to identify five landscape indicators: amount, distribution, patch size, structural connectivity, and diversity of vegetation types. We then conducted pilot applications in four bureau of land management (BLM) field offices in Arizona, California, and Colorado to refine procedures for quantifying the indicators and assess the utility of the indicators for managers. Results highlighted the dominance of upland and the limited extent of riparian/wetland vegetation communities, moderate connectivity of priority vegetation patches, and lower diversity of native vegetation types on BLM compared to non-BLM lands. Agency staff can use the indicators to inform the development of quantitative resource management objectives in land use plans, evaluate progress in meeting those objectives, quantify potential impacts of proposed actions, and as a foundation for an all-lands approach to landscape-level management across public lands.

Environmental Management↗
USGS2025

“Leaky weirs” capture alluvial deposition and enhance seasonal mountain-front recharge in dryland streams

“Leaky weirs” are rock structures installed in dryland streams, which are anchored into exposed bedrock, loosely cemented, and designed to allow water to slowly pass through. They are being tested at a ranch in southeastern Arizona, USA, to restore and conserve the historic range and desert wetlands. Data are collected to assess how leaky weirs impact surface water, subsurface water, and groundwater recharge—including stream discharge, timing, and depth of infiltration, and groundwater elevations. Three adjacent watersheds, two with outlets just below leaky weirs and one with leaky weirs farther upstream, were instrumented with water-level loggers, wildlife cameras, and crest stage instruments with temperature sensors in the soil. As most groundwater recharge is assumed to be focused along the mountain fronts in this region, mountain-block recharge is also evaluated to differentiate between the two using isotope analyses. Finally, a single, late-season flood event is scrutinized to consider the leaky weir effect on all monitored components in the water budget. Results indicated groundwater flow is primarily from the mountains to the east via older, regional mountain-block recharge. However, the development of shallow alluvial aquifers is supported by the leaky weirs, that slow flows, capture permeable sediments, and allow infiltration, thus enhancing mountain-front recharge. In turn, these new pockets of water help support the restoration of historic wetlands. Sediment accumulates where leaky weirs are installed, reducing flashy peak flows, and resulting in a series of infiltration ponds along the channel that support vegetation during growing seasons and recharge the shallow aquifer during non-growing seasons.

Arizona↗
USGS2002

Assemblages of breeding birds as indicators of grassland condition

We developed a measure of biological integrity for grasslands (GI) based on the most influential habitat types in the Prairie Pothole Region of North Dakota. GI is based on proportions of habitat types and the relationships of these habitat types to breeding birds. Habitat types were identified by digital aerial photography, verified on the ground, and quantified using GIS, We then developed an index to GI based on the presence or abundance of breeding bird species. Species abundance data were obtained from 3 min roadside point counts at 889 points in 44, 4050 ha study plots over a 2-year period. Using a modified North American Breeding Bird Survey protocol, species were recorded in each of four quadrants at each point. Fifty species selected for analysis included all grassland species that occurred in at least 15 quadrants and all other bird species that occurred in at least 1 % of quadrants. We constructed preliminary models using data from each of the 2 years, then tested their predictive ability by cross-validation with data from the other year. These cross-validation tests indicated that the index consistently predicted grassland integrity. The final four models (presence and abundance models at 200 and 400 m scales) included only those species that were statistically significant (P ??? 0.05) in all preliminary models. Finally, we interpreted the components of the indices by examining associations between individual species and habitat types. Logistic regression identified 386 statistically significant relationships between species and habitat types at 200 and 400m scales. This method, though labor-intensive, successfully uses the presence of grassland-dependent species and absence of species associated with woody vegetation or cropland to provide an index to grassland integrity. Once regional associations of species with habitat types have been identified, such indices can be applied relatively inexpensively to monitor grassland integrity over large geographic areas. Indices like the ones presented here could be applied widely using bird abundance data that are currently being collected across the United States and southern Canada through the North American Breeding Bird Survey. ?? 2002 Elsevier Science Ltd. All rights reserved.

Ecological Indicators↗
USGS2011

Adapting to climate change at Olympic National Forest and Olympic National Park

Climate change presents a major challenge to natural resource managers both because of the magnitude of potential effects of climate change on ecosystem structure, processes, and function, and because of the uncertainty associated with those potential ecological effects. Concrete ways to adapt to climate change are needed to help natural resource managers take the first steps to incorporate climate change into management and take advantage of opportunities to counteract the negative effects of climate change. We began a climate change adaptation case study at Olympic National Forest (ONF) in partnership with Olympic National Park (ONP) to determine how to adapt management of federal lands on the Olympic Peninsula, Washington, to climate change. The case study began in the summer of 2008 and continued for 1½ years. The case study process involved science-based sensitivity assessments, review of management activities and constraints, and adaptation workshops in each of four focus areas (hydrology and roads, fish, vegetation, and wildlife). The process produced adaptation options for ONF and ONP, and illustrated the utility of place-based vulnerability assessment and science-management workshops in adapting to climate change. The case study process provides an example for other national forests, national parks, and natural resource agencies of how federal land management units can collaborate in the initial stages of climate change adaptation. Many of the ideas generated through this process can potentially be applied in other locations and in other agencies

Washington↗
USGS2015

Hydrology: The interdisciplinary science of water

We live in a world where biophysical and social processes are tightly coupled. Hydrologic systems change in response to a variety of natural and human forces such as climate variability and change, water use and water infrastructure, and land cover change. In turn, changes in hydrologic systems impact socioeconomic, ecological, and climate systems at a number of scales, leading to a coevolution of these interlinked systems. The Harvard Water Program, Hydrosociology, Integrated Water Resources Management, Ecohydrology, Hydromorphology, and Sociohydrology were all introduced to provide distinct, interdisciplinary perspectives on water problems to address the contemporary dynamics of human interaction with the hydrosphere and the evolution of the Earth’s hydrologic systems. Each of them addresses scientific, social, and engineering challenges related to how humans influence water systems and vice versa. There are now numerous examples in the literature of how holistic approaches can provide a structure and vision of the future of hydrology. We review selected examples, which taken together, describe the type of theoretical and applied integrated hydrologic analyses and associated curricular content required to address the societal issue of water resources sustainability. We describe a modern interdisciplinary science of hydrology needed to develop an in-depth understanding of the dynamics of the connectedness between human and natural systems and to determine effective solutions to resolve the complex water problems that the world faces today. Nearly, every theoretical hydrologic model introduced previously is in need of revision to accommodate how climate, land, vegetation, and socioeconomic factors interact, change, and evolve over time.

Water Resources Research↗
USGS2020

Vegetation sampling and management

What is the utility of vegetation measurements for wildlife managers? In the prairie, savanna, tundra, forest, steppe, and wetland regions of the world, mixtures of plant species provide wildlife with food, cover and, in some circumstances, water; the 3 essential habitat elements necessary to sustain viable wildlife populations. We define habitat in reference to use of a vegetation type by an animal (e.g., deer habitat) and vegetation type when referring to differences in vegetation stands (e.g., marsh vegetation type versus tall grass prairie vegetation type; Hall et al. 1997). In strict definition, the variety of wildlife using plants ranges from snails and voles (Microtus spp.) to bison (Bison bison) and elephants (Loxodonta spp.) in uplands and from mosquitoes and ducks to muskrats (Ondatra zibethicus) and manatees (Trichechus manatus) in wetlands. Through evolutionary processes, some wildlife species are totally dependent on vegetation for all annual life requirements, whereas other species use vegetation only for cover or food. Regardless of the role of vegetation in the sustenance of wildlife, any management or research project that requires evaluation of wildlife and vegetation type relationships on a unit of land will necessitate some form of vegetation measurement. The term vegetation can refer to a single plant or species on a specific site or a community in the landscape. Vegetation may occur naturally or be introduced, and may be live or dead. Uses of vegetation measurements are many: (1) evaluation of vegetation response to management practices, (2) estimation of carrying capacity and/or forage production, (3) characterization of cover and habitat components for an endangered species, or (4) long-term monitoring of the general trend of plant vigor or vegetation type condition. Surveying and measuring quantity and quality of vegetation within habitats are basic to wildlife research and management. Grassland, shrubland, and woodland vegetation types are comprised of populations in which individual plants are usually too numerous to inventory completely. Consequently, wildlife biologists usually use sampling techniques to make inferences about the total plant population within a given vegetation type. Vegetation sampling methodologies have evolved within several ecological disciplines (e.g., plant ecology, forestry, rangeland science) and for a variety of management or research objectives (e.g., estimating forage for ungulates, describing habitat use by passerine birds). Description of every method that has been used to sample vegetation is beyond the scope of this chapter. We describe how to measure vegetation structure, which Dansereau (1957) defined as the spatial organization (distribution) of individuals that form a stand. We have organized this chapter into a description of basic methods of vegetation sampling with examples of how those methods have been applied or modified in wildlife research and management. We assume the investigator/reader has adequate knowledge of the concepts of wildlife ecology, primary habitat requirements of wildlife species under study, and ability to systematically identify the species of wildlife and vascular plants within the geographical area of investigation.

Book chapter↗
USGS2010

Conifer health classification for Colorado, 2008

Colorado has undergone substantial changes in forests due to urbanization, wildfires, insect-caused tree mortality, and other human and environmental factors. The U.S. Geological Survey Rocky Mountain Geographic Science Center evaluated and developed a methodology for applying remotely-sensed imagery for assessing conifer health in Colorado. Two classes were identified for the purposes of this study: healthy and unhealthy (for example, an area the size of a 30- x 30-m pixel with 20 percent or greater visibly dead trees was defined as ?unhealthy?). Medium-resolution Landsat 5 Thematic Mapper imagery were collected. The normalized, reflectance-converted, cloud-filled Landsat scenes were merged to form a statewide image mosaic, and a Normalized Difference Vegetation Index (NDVI) and Renormalized Difference Infrared Index (RDII) were derived. A supervised maximum likelihood classification was done using the Landsat multispectral bands, the NDVI, the RDII, and 30-m U.S. Geological Survey National Elevation Dataset (NED). The classification was constrained to pixels identified in the updated landcover dataset as coniferous or mixed coniferous/deciduous vegetation. The statewide results were merged with a separate health assessment of Grand County, Colo., produced in late 2008. Sampling and validation was done by collecting field data and high-resolution imagery. The 86 percent overall classification accuracy attained in this study suggests that the data and methods used successfully characterized conifer conditions within Colorado. Although forest conditions for Lodgepole Pine (Pinus contorta) are easily characterized, classification uncertainty exists between healthy/unhealthy Ponderosa Pine (Pinus ponderosa), Pi?on (Pinus edulis), and Juniper (Juniperus sp.) vegetation. Some underestimation of conifer mortality in Summit County is likely, where recent (2008) cloud-free imagery was unavailable. These classification uncertainties are primarily due to the spatial and temporal resolution of Landsat, and of the NLCD derived from this sensor. It is believed that high- to moderate-resolution multispectral imagery, coupled with field data, could significantly reduce the uncertainty rates. The USGS produced a four-county follow-up conifer health assessment using high-resolution RapidEye remotely sensed imagery and field data collected in 2009.

Scientific Investigations Map↗
USGS2001

Pollen assemblages as paleoenvironmental proxies in the Florida Everglades

Analysis of 170 pollen assemblages from surface samples in eight vegetation types in the Florida Everglades indicates that these wetland sub-environments are distinguishable from the pollen record and that they are useful proxies for hydrologic and edaphic parameters. Vegetation types sampled include sawgrass marshes, cattail marshes, sloughs with floating aquatics, wet prairies, brackish marshes, tree islands, cypress swamps, and mangrove forests. The distribution of these vegetation types is controlled by specific environmental parameters, such as hydrologic regime, nutrient availability, disturbance level, substrate type, and salinity; ecotones between vegetation types may be sharp. Using R-mode cluster analysis of pollen data, we identified diagnostic species groupings; Q-mode cluster analysis was used to differentiate pollen signatures of each vegetation type. Cluster analysis and the modern analog technique were applied to interpret vegetational and environmental trends over the last two millennia at a site in Water Conservation Area 3A. The results show that close modern analogs exist for assemblages in the core and indicate past hydrologic changes at the site, correlated with both climatic and land-use changes. The ability to differentiate marshes with different hydrologic and edaphic requirements using the pollen record facilitates assessment of relative impacts of climatic and anthropogenic changes on this wetland ecosystem on smaller spatial and temporal scales than previously were possible. ?? 2001 Elsevier Science B.V.

Review of Palaeobotany and Palynology↗
USGS2017

NDVI, scale invariance and the modifiable areal unit problem: An assessment of vegetation in the Adelaide Parklands

This research addresses the question as to whether or not the Normalised Difference Vegetation Index (NDVI) is scale invariant (i.e. constant over spatial aggregation) for pure pixels of urban vegetation. It has been long recognized that there are issues related to the modifiable areal unit problem (MAUP) pertaining to indices such as NDVI and images at varying spatial resolutions. These issues are relevant to using NDVI values in spatial analyses. We compare two different methods of calculation of a mean NDVI: 1) using pixel values of NDVI within feature/object boundaries and 2) first calculating the mean red and mean near-infrared across all feature pixels and then calculating NDVI. We explore the nature and magnitude of these differences for images taken from two sensors, a 1.24 m resolution WorldView-3 and a 0.1 m resolution digital aerial image. We apply these methods over an urban park located in the Adelaide Parklands of South Australia. We demonstrate that the MAUP is not an issue for calculation of NDVI within a sensor for pure urban vegetation pixels. This may prove useful for future rule-based monitoring of the ecosystem functioning of green infrastructure.

Adelaide Parklands↗
USGS2010

Establishing a Multi-scale Stream Gaging Network in the Whitewater River Basin, Kansas, USA

Investigating the routing of streamflow through a large drainage basin requires the determination of discharge at numerous locations in the channel network. Establishing a dense network of stream gages using conventional methods is both cost-prohibitive and functionally impractical for many research projects. We employ herein a previously tested, fluid-mechanically based model for generating rating curves to establish a stream gaging network in the Whitewater River basin in south-central Kansas. The model was developed for the type of channels typically found in this watershed, meaning that it is designed to handle deep, narrow geomorphically stable channels with irregular planforms, and can model overbank flow over a vegetated floodplain. We applied the model to ten previously ungaged stream reaches in the basin, ranging from third- to sixth-order channels. At each site, detailed field measurements of the channel and floodplain morphology, bed and bank roughness, and vegetation characteristics were used to quantify the roughness for a range of flow stages, from low flow to overbank flooding. Rating curves that relate stage to discharge were developed for all ten sites. Both fieldwork and modeling were completed in less than 2 years during an anomalously dry period in the region, which underscores an advantage of using theoretically based (as opposed to empirically based) discharge estimation techniques. ?? 2010 Springer Science+Business Media B.V.

Water Resources Management↗
USGS2014

Hydraulic modeling development and application in water resources engineering

The use of modeling has become widespread in water resources engineering and science to study rivers, lakes, estuaries, and coastal regions. For example, computer models are commonly used to forecast anthropogenic effects on the environment, and to help provide advanced mitigation measures against catastrophic events such as natural and dam-break floods. Linking hydraulic models to vegetation and habitat models has expanded their use in multidisciplinary applications to the riparian corridor. Implementation of these models in software packages on personal desktop computers has made them accessible to the general engineering community, and their use has been popularized by the need of minimal training due to intuitive graphical user interface front ends. Models are, however, complex and nontrivial, to the extent that even common terminology is sometimes ambiguous and often applied incorrectly. In fact, many efforts are currently under way in order to standardize terminology and offer guidelines for good practice, but none has yet reached unanimous acceptance. This chapter provides a view of the elements involved in modeling surface flows for the application in environmental water resources engineering. It presents the concepts and steps necessary for rational model development and use by starting with the exploration of the ideas involved in defining a model. Tangible form of those ideas is provided by the development of a mathematical and corresponding numerical hydraulic model, which is given with a substantial amount of detail. The issues of model deployment in a practical and productive work environment are also addressed. The chapter ends by presenting a few model applications highlighting the need for good quality control in model validation.

Book chapter↗
USGS2011

Structural and functional effects of herbicides on non-target organisms in aquatic ecosystems with an emphasis on atrazine

Herbicide use has increased dramatically around the world over the past 6 decades (Gianessi and Reigner, 2007). Few herbicides were in use in the 1950s. However, by 2001 approximately 1.14 billion kilograms of herbicides were applied globally for the control of undesireable vegetation in agricultural, silvicultural, lawncare, aquacultural, and irrigation/recreational water management activities (Kiely et al., 2004). Twenty-eight percent of the total mass of herbicides is applied in the United States, with the remaining 72 percent being applied elsewhere around the globe (Kiely et al., 2004). Herbicides represent 36% of global pesticide use, followed by insecticides (25%), fungicides (10%) and other chemical classes (Kiely et al., 2004). Agricultural production accounts for approximately 90% of herbicide use in the U.S. (Kiely et al., 2004). Gianessi and Reigner (2007) indicated that herbicides are routinely used on more than 90% of the area designated for large commercial crops including corn, soybeans, cotton, sugar beets, peanuts, and rice. Increased farm mechanization, technological advancements in production of inexpensive sources of inorganic nitrogen fertilizer (e.g., anhydrous ammonia), and conversion of forest, grassland, and wetland habitats to cropland has led to a tremendous increase in global food production over the past half-century. Herbicides have augmented advances in large-scale agricultural systems and have largely replaced mechanical and hand-weeding control mechanisms (Gianessi and Reigner, 2007). The wide-spread use of herbicides in agriculture has resulted in frequent chemical detections in surface and groundwaters (Gilliom, 2007). The majority of herbicides used are highly water soluble and are therefore prone to runoff from terrestrial environments. In additon, spray drift and atmospheric deposition can contribute to herbicide contamination of aquatic environments. Lastly, selected herbicides are deliberately applied to aquatic environments for controlling nuisance aquatic vegetation. Although aquatic herbicide exposure has been widely documented, these exposures are not necessarily related to adverse non-target ecological effects on natural communities in aquatic environments. This chapter evaluates the potential for effects of herbicides on the structure and function of aquatic envrionments at the population, community, and ecosystem levels of biological organization. In this manuscript I examine several critical aspects of the subject matter area: primary herbicides in use and chemical modes of action; the regulatory process used for registration and risk assessment of herbicides; data regarding non-target risks and the relative sensitivity of aquatic plants, inveretebrates, and fish to herbicides; and emerging areas of science regarding the potential for endocrine-disrupting effects of herbicides on aquatic vertebrates. Much of the focus of this paper is on atrazine due to the extensive database which exists regarding its fate and effects.

Book chapter↗
USGS2016

Modeling the effects of land cover and use on landscape capability for urban ungulate populations

Expanding ungulate populations are causing concerns for wildlife professionals and residents in many urban areas worldwide. Nowhere is the phenomenon more apparent than in the eastern US, where urban white-tailed deer (Odocoileus virginianus) populations are increasing. Most habitat suitability models for deer have been developed in rural areas and across large (>1000 km2) spatial extents. Only recently have we begun to understand the factors that contribute to space use by deer over much smaller spatial extents. In this study, we explore the concepts, terminology, methodology and state-of-the-science in wildlife abundance modeling as applied to overabundant deer populations across heterogeneous urban landscapes. We used classified, high-resolution digital orthoimagery to extract landscape characteristics in several urban areas of upstate New York. In addition, we assessed deer abundance and distribution in 1-km2 blocks across each study area from either aerial surveys or ground-based distance sampling. We recorded the number of detections in each block and used binomial mixture models to explore important relationships between abundance and key landscape features. Finally, we cross-validated statistical models of abundance and compared covariate relationships across study sites. Study areas were characterized along a gradient of urbanization based on the proportions of impervious surfaces and natural vegetation which, based on the best-supported models, also distinguished blocks potentially occupied by deer. Models performed better at identifying occurrence of deer and worse at predicting abundance in cross-validation comparisons. We attribute poor predictive performance to differences in deer population trajectories over time. The proportion of impervious surfaces often yielded better predictions of abundance and occurrence than did the proportion of natural vegetation, which we attribute to a lack of certain land cover classes during cold and snowy winters. Merits and limitations of our approach to habitat suitability modeling are discussed in detail.

Book chapter↗
USGS2021

Techniques to improve ecological interpretability of black box machine learning models

Statistical modeling of ecological data is often faced with a large number of variables as well as possible nonlinear relationships and higher-order interaction effects. Gradient boosted trees (GBT) have been successful in addressing these issues and have shown a good predictive performance in modeling nonlinear relationships, in particular in classification settings with a categorical response variable. They also tend to be robust against outliers. However, their black-box nature makes it difficult to interpret these models. We introduce several recently developed statistical tools to the environmental research community in order to advance interpretation of these black-box models. To analyze the properties of the tools, we applied gradient boosted trees to investigate biological health of streams within the contiguous USA, as measured by a benthic macroinvertebrate biotic index. Based on these data and a simulation study, we demonstrate the advantages and limitations of partial dependence plots (PDP), individual conditional expectation (ICE) curves and accumulated local effects (ALE) in their ability to identify covariate–response relationships. Additionally, interaction effects were quantified according to interaction strength (IAS) and Friedman’s H 2 "> H 2 statistic. Interpretable machine learning techniques are useful tools to open the black-box of gradient boosted trees in the environmental sciences. This finding is supported by our case study on the effect of impervious surface on the benthic condition, which agrees with previous results in the literature. Overall, the most important variables were ecoregion, bed stability, watershed area, riparian vegetation and catchment slope. These variables were also present in most identified interaction effects. In conclusion, graphical tools (PDP, ICE, ALE) enable visualization and easier interpretation of GBT but should be supported by analytical statistical measures. Future methodological research is needed to investigate the properties of interaction tests. Supplementary materials accompanying this paper appear on-line.

Journal of Agricultural, Biological, and Environme↗
USGS2023

Modeling post-wildfire hydrologic response: Review and future directions for applications of physically based distributed simulation

Wildfire is a growing concern as climate shifts. The hydrologic effects of wildfire, which include elevated hazards and changes in water quantity and quality, are increasingly assessed using numerical models. Post-wildfire application of physically based distributed models provides unique insight into the underlying processes that affect water resources after wildfire. This work reviews and synthesizes post-wildfire applications of physically based distributed models by examining the scales and geographic/ecohydrologic distribution of model applications, hydrologic response process representation, model parameterization, and model performance metrics. Highlighted gaps and opportunities for advancing physically based distributed hydrologic response modeling after wildfire include the following: (a) applying models in under-represented geographic (S. America, Africa, Asia) and ecohydrologic regions (arid or dry subhumid climates), (b) incorporating all four major streamflow generation mechanisms (infiltration excess, saturation excess, subsurface storm flow, and groundwater flow), (c) representing integrated vadose zone and saturated zone processes to better capture subsurface streamflow generation, (d) building new remotely sensed model parameterization methods for precipitation interception, infiltration, and overland flow that account for burn severity and recovery, (e) incorporating distributed state variables (e.g., soil moisture, groundwater levels) in model performance assessment, (f) designing model intercomparison studies, including field datasets specifically for post-wildfire model development and validation, (g) linking mechanistic vegetation regrowth models with hydrologic models to improve simulation of process shifts as ecosystems recover, and (h) creating a new community modeling framework to integrate modeling advances across the wildfire science community.

Earth's Future↗
USGS2025

Potential risks of vegetation treatments to introduce and increase invasive annual grasses in rangelands of Montana, North Dakota, and South Dakota— A science synthesis to inform National Environmental Policy Act analyses

The U.S. Geological Survey is working with Federal land management agencies to develop a series of planned structured science syntheses to support environmental effects analyses that agencies conduct under the National Environmental Policy Act (NEPA). This report synthesizes science information relevant to environmental effects analyses concerned with potential increases in the distribution and abundance of invasive annual grasses (IAGs) from proposed vegetation treatments for habitat restoration. The focal environments for this synthesis are rangelands in the intermontane valleys of Montana and the northern Great Plains of Montana, North Dakota, and South Dakota. The synthesis is organized to align with the standard elements of NEPA analyses and provides information on relevant scientific studies, data availability, analysis methods, and mitigation measures. We found that the likelihood of increasing IAGs from vegetation treatments depends on treatment type and environmental context. In sagebrush ecosystems of the focal region, prescribed fire often reduces or does not increase IAGs. Treatments that cause soil disturbances, such as mechanical removals of sagebrush or firebreak constructions, are more likely to increase IAGs than other treatments. Herbicides applied to reduce sagebrush cover have not increased the proportion of IAGs in the plant community. Temperature and precipitation have been strong factors in determining IAG responses to vegetation treatments in sagebrush ecosystems of the focal region, where more precipitation in spring and summer likely provides a competitive edge to native, perennial grasses more than winter annual grasses like Bromus tectorum L. (cheatgrass). In grasslands, prescribed fire often reduces IAGs, but effects depend on the abundance of native species and are often short lived. Mowing can increase or decrease IAGs in grassland ecosystems. Grassland site conditions, such as southeast-facing slopes, sandier or rockier sites, or lower native species cover or richness affect the likelihood of invasion by annual grasses. Maintaining adequate cover of perennial vegetation creates rangelands that are resistant and resilient to annual grass invasions. Managers can minimize invasion potential by focusing on treatment type, placement, and seasonal timing. Herbicides also can provide effective mitigation, especially in combination with other controls such as prescribed fire or grazing. This report can be incorporated by reference in NEPA documentation, included in a project record, or provide a general reference for understanding and identifying literature about increases in IAGs associated with vegetation treatments in rangelands in this focal region.

Montana, North Dakota, South Dakota↗