USGS Science⌕ Search

SEARCH · USGS Science

Results for “Modeling and Using Context”

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 559 records · Page 31Linked to original sources
USGS2018

Simulation of hydrodynamics, water quality, and lake sturgeon habitat volumes in Lake St. Croix, Wisconsin and Minnesota, 2013

Lake St. Croix is a naturally impounded, riverine lake that makes up the last 40 kilometers of the St. Croix River. Substantial land-use changes during the past 150 years, including increased agriculture and urban development, have reduced Lake St. Croix water-quality and increased nutrient loads delivered to Lake St. Croix. A recent (2012–13) total maximum daily load phosphorus-reduction plan set the goal to reduce total phosphorus loads to Lake St. Croix by 20 percent by 2020 and reduce Lake St. Croix algal bloom frequencies. The U.S. Geological Survey, in cooperation with the National Park Service, developed a two-dimensional, carbon-based, laterally averaged, hydrodynamic and water-quality model, CE–QUAL–W2, that addresses the interaction between nutrient cycling, primary production, and trophic dynamics to predict responses in the distribution of water temperature, oxygen, and chlorophyll a. Distribution is evaluated in the context of habitat for lake sturgeon, including a combination of temperature and dissolved oxygen conditions termed oxy-thermal habitat. The Lake St. Croix CE–QUAL–W2 model successfully reproduced temperature and dissolved oxygen in the lake longitudinally (from upstream to downstream), vertically, and temporally over the seasons. The simulated water temperature profiles closely matched the measured water temperature profiles throughout the year, including the prediction of thermocline transition depths (often within 1 meter), the absolute temperature of the thermocline transitions (often within 1.0 degree Celsius), and profiles without a strong thermocline transition. Simulated dissolved oxygen profiles matched the trajectories of the measured dissolved oxygen concentrations at multiple depths over time, and the simulated concentrations matched the depth and slope of the measured concentrations. Additionally, trends in the measured water-quality data were captured by the model simulation, gaining some potential insights into the underlying mechanisms of critical Lake St. Croix metabolic processes. The CE–QUAL–W2 model tracked nitrate plus nitrite, total nitrogen, and total phosphorus throughout the year. Inflow nutrient contributions (loads), largely dominated by upstream St. Croix River loads, were the most important controls on Lake St. Croix water quality. Close to 60 percent of total phosphorus to the lake was from phosphorus derived from organic matter, and about 89 percent of phosphorus to Lake St. Croix was delivered by St. Croix River inflows. The Lake St. Croix CE–QUAL–W2 model offered potential mechanisms for the effect of external and internal loadings on the biotic response regarding the modeled algal community types of diatoms, green algae, and blue-green algae. The model also suggested the seasonal dominance of blue-green algae in all four pools of the lake. A sensitivity analysis was completed to test the total maximum daily load phosphorus-reduction scenario responses of total phosphorus and chlorophyll a. The modeling indicates that phosphorus reductions would result in similar Lake St. Croix reduced concentrations, although chlorophyll a concentrations did not decrease in the same proportional amounts as the total phosphorus concentrations had decreased. The smaller than expected reduction in algal growth rates highlighted that although inflow phosphorus loads are important, other constituents also can affect the algal response of the lake, such as changes in light penetration and the breakdown of organic matter releasing nutrients. The available habitat suitable for lake sturgeon was evaluated using the modeling results to determine the total volume of good-growth habitat, optimal growth habitat, and lethal temperature habitat. Overall, with the calibrated model, the fish habitat volume in general contained a large proportion of good-growth habitat and a sustained period of optimal growth habitat in the summer. Only brief periods of lethal oxy-thermal habitat were present in Lake St. Croix during the model simulation.

Minnesota, Wisconsin↗
USGS2025

Greater sage-grouse seasonal habitat associations: A review and considerations for interpretation and management applications

Habitat features needed by wildlife can change in composition throughout the year, particularly in temperate ecosystems, leading to distinct seasonal spatial-use patterns. Studies of species-habitat associations therefore often focus on understanding relationships within discrete seasonal periods with common goals of prediction (e.g., habitat mapping) and inference (e.g., interpreting model coefficients). Across the range of the greater sage-grouse ( Centrocercus urophasianus ) of western North America, the increasing use of high-frequency tracking devices has led to a surge in habitat association studies covering multiple temporal periods and spatial extents. We reviewed the literature for seasonal habitat association studies corresponding to the second and third orders of selection (Johnson 1980). Our objectives were to summarize the methodological approaches used to estimate habitat associations to aid understanding in cross-study comparisons and identify common habitat features reported as selected or avoided within different seasonal periods. We reviewed 71 second- and third-order studies published from 2007–2023 that evaluated covariates collected in a geographic information system (GIS) and modeled probability of selection or intensity of use. The most common study design evaluated a single level of selection (third-order) and was multi-scale (i.e., covariates were measured at varying spatial grains). The most common model approach estimated habitat associations using resource selection functions (RSFs) fit with logistic regression. Studies mostly focused on the breeding periods and winter, but all seasons throughout the annual cycle were covered. There was clear support for selection of sagebrush and avoidance of trees and rugged terrain across seasons, and strong selection of mesic conditions in summer. However, habitat associations for most covariates were mixed, with proportionally equivalent selection and avoidance reported, even within the same seasons. Different factors hampered cross-study comparisons, including variation in study design, but additional contributors likely included important context-dependent habitat associations, such as functional responses to changing habitat availability. We suggest collaborative studies leveraging multiple datasets can help improve seasonal habitat inference by removing the effects of variable study designs.

Journal of Wildlife Management↗
USGS2023

Linking vocal behaviours to habitat structure to create behavioural landscapes

The recent development of animal-borne sensors coupled with location data can provide insights into how individuals modify their behaviour with respect to specific habitat features. Animals can express a diverse array of behaviours as they navigate heterogenous landscapes, yet few studies have specifically evaluated the interaction of behaviours with habitat characteristics. We used a novel broadcast acoustic transmitter to investigate the interaction between vocal behaviours of an endemic Hawaiian thrush, the ʻōmaʻo, Myadestes obscurus , and habitat features across a naturally fragmented forest landscape. Through the development of behavioural landscape models that link specific vocalizations with space use, we found that the use of different vocalization types (calls, songs, whisper songs) were highly variable across the landscape but were associated with distinct habitat features. The likelihood of calls increased in an open lava matrix between forest patches, while whisper songs were more strongly associated with the dense interior areas of forest fragments. In contrast, the rate of ʻōmaʻo vocalizations overall decreased in the open lava matrix, suggesting that ʻōmaʻo may shift behaviours from territory defence to foraging as they transition through different habitats. Our study revealed context-specific changes in behaviour across ʻōmaʻo home ranges, including courtship, aggression and social interactions between individuals. Combining the use of a novel acoustic tool with automated radiotelemetry allowed us to overcome challenges associated with detection and analysis of variation in behaviour and resource selection across a highly heterogeneous landscape that would have been otherwise difficult to impossible.

Hawaii↗
USGS2019

Multiorder hydrologic position in the conterminous United States: A set of metrics in support of groundwater mapping at regional and national scales

The location of a point on the landscape within a stream network (hydrologic position) can be an important predictive measure in hydrology. Hydrologic position is defined here by two metrics: lateral position and distance from stream to divide, both measured horizontally. Lateral position (dimensionless) is the relative position of a point between the stream and its watershed divide. Distance from stream to divide (units of length) is an indicator of position within a watershed: generally small near a confluence and generally large in headwater areas. Watersheds and watershed divides are defined here by Thiessen polygons rather than topographic divides. Lateral position and distance from stream to divide are also defined in the context of hydrologic order. Hydrologic order “ n ” is defined as the network of streams, and associated divides, of order n and higher. And given that a point can have different positions in different hydrologic orders the term multiorder hydrologic position (MOHP) is used to describe the ensemble of hydrologic positions. MOHP was mapped across the conterminous United States for nine hydrologic orders at a spatial resolution of 30 m (about 8.7 billion pixels). There are 18 metrics for each pixel. Four case studies are presented that use MOHP metrics as explanatory factors in random forest machine learning models. The case studies show that lower order MOHP metrics can serve as indicators of hydrologic process while higher‐order metrics serve as indicators of location. MOHP is shown to have utility as a predictor variable across a large range of scales (50,000 to 8,000,000 km 2 ).

Water Resources Research↗
USGS2014

Continuing megathrust earthquake potential in Chile after the 2014 Iquique earthquake

The seismic gap theory identifies regions of elevated hazard based on a lack of recent seismicity in comparison with other portions of a fault. It has successfully explained past earthquakes (see, for example, ref. 2) and is useful for qualitatively describing where large earthquakes might occur. A large earthquake had been expected in the subduction zone adjacent to northern Chile which had not ruptured in a megathrust earthquake since a M ~8.8 event in 1877. On 1 April 2014 a M 8.2 earthquake occurred within this seismic gap. Here we present an assessment of the seismotectonics of the March–April 2014 Iquique sequence, including analyses of earthquake relocations, moment tensors, finite fault models, moment deficit calculations and cumulative Coulomb stress transfer. This ensemble of information allows us to place the sequence within the context of regional seismicity and to identify areas of remaining and/or elevated hazard. Our results constrain the size and spatial extent of rupture, and indicate that this was not the earthquake that had been anticipated. Significant sections of the northern Chile subduction zone have not ruptured in almost 150 years, so it is likely that future megathrust earthquakes will occur to the south and potentially to the north of the 2014 Iquique sequence.

Nature↗
USGS2011

Large-scale flow experiments for managing river systems

Experimental manipulations of streamflow have been used globally in recent decades to mitigate the impacts of dam operations on river systems. Rivers are challenging subjects for experimentation, because they are open systems that cannot be isolated from their social context. We identify principles to address the challenges of conducting effective large-scale flow experiments. Flow experiments have both scientific and social value when they help to resolve specific questions about the ecological action of flow with a clear nexus to water policies and decisions. Water managers must integrate new information into operating policies for large-scale experiments to be effective. Modeling and monitoring can be integrated with experiments to analyze long-term ecological responses. Experimental design should include spatially extensive observations and well-defined, repeated treatments. Large-scale flow manipulations are only a part of dam operations that affect river systems. Scientists can ensure that experimental manipulations continue to be a valuable approach for the scientifically based management of river systems.

BioScience↗
USGS2017

Landscape complementation revealed through bipartite networks: An example with the Florida manatee

Context Landscape complementation, or how landscapes that contain two or more non-substitutable and spatially separated resources facilitate resource use, is critical for many populations. Implicit to the problem of landscape complementation is the movement of individuals to access multiple resources. Conventional measures of complementation, such as habitat area or distance between habitats, do not consider the spatial configuration of resources or how landscape features impede movement. Objectives We advanced a bipartite network approach to capture the spatial configuration and connectivity of two habitat types and contrasted this framework to conventional approaches in a habitat selection model. Methods Using satellite-telemetry of the Florida manatee (Trichechus manatus latirostris), a marine mammal that relies on two distinct, spatially separate habitats for foraging and thermoregulating, we parameterized and compared mixed conditional logistic models with covariates describing classic habitat selection metrics, conventional measures of landscape complementation, and bipartite network metrics. Results The models best supported included habitat area, resistance distance between habitats, and the bipartite network metric eigenvector centrality. The connectivity between habitats and the spatial configuration of one habitat type relative to other types better described habitat selection than conventional measures of landscape complementation alone. The type of habitat, i.e. seagrass or thermal refuge, influenced both the direction and magnitude of the response. Conclusions Landscape complementation is an important predictor of selection and thus classic complementation measures are not sufficient in describing the process. Formalization of complementation with bipartite network can therefor reveal effects potentially missed with conventional measures.

Florida↗
USGS2005

Geophysical monitoring of a field-scale biostimulation pilot project

The USGS conducted a geophysical investigation in support of a U.S. Naval Facilities Engineering Command, Southern Division field‐scale biostimulation pilot project at Anoka County Riverfront Park (ACP), downgradient of the Naval Industrial Reserve Ordnance Plant, Fridley, Minnesota. The goal of the pilot project is to evaluate subsurface injection of vegetable oil emulsion (VOE) to stimulate microbial degradation of chlorinated hydrocarbons. To monitor the emplacement and movement of the VOE and changes in water chemistry resulting from VOE dissolution and/or enhanced biological activity, the USGS acquired cross‐hole radar zero‐offset profiles, traveltime tomograms, and borehole geophysical logs during five site visits over 1.5 years. Analysis of pre‐ and postinjection data sets using petrophysical models developed to estimate VOE saturation and changes in total dissolved solids provides insights into the spatial and temporal distribution of VOE and ground water with altered chemistry. Radar slowness‐difference tomograms and zero‐offset slowness profiles indicate that the VOE remained close to the injection wells, whereas radar attenuation profiles and electromagnetic induction logs indicate that bulk electrical conductivity increased downgradient of the injection zone, diagnostic of changing water chemistry. Geophysical logs indicate that some screened intervals were located above or below zones of elevated dissolved solids; hence, the geophysical data provide a broader context for interpretation of water samples and evaluation of the biostimulation effort. Our results include (1) demonstration of field and data analysis methods for geophysical monitoring of VOE biostimulation and (2) site‐specific insights into the spatial and temporal distributions of VOE at the ACP.

Minnesota↗
USGS2010

Linking microbial and ecosystem ecology using ecological stoichiometry: a synthesis of conceptual and empirical approaches

Currently, one of the biggest challenges in microbial and ecosystem ecology is to develop conceptual models that organize the growing body of information on environmental microbiology into a clear mechanistic framework with a direct link to ecosystem processes. Doing so will enable development of testable hypotheses to better direct future research and increase understanding of key constraints on biogeochemical networks. Although the understanding of phenotypic and genotypic diversity of microorganisms in the environment is rapidly accumulating, how controls on microbial physiology ultimately affect biogeochemical fluxes remains poorly understood. We propose that insight into constraints on biogeochemical cycles can be achieved by a more rigorous evaluation of microbial community biomass composition within the context of ecological stoichiometry. Multiple recent studies have pointed to microbial biomass stoichiometry as an important determinant of when microorganisms retain or recycle mineral nutrients. We identify the relevant cellular components that most likely drive changes in microbial biomass stoichiometry by defining a conceptual model rooted in ecological stoichiometry. More importantly, we show how X-ray microanalysis (XRMA), nanoscale secondary ion mass spectroscopy (NanoSIMS), Raman microspectroscopy, and in situ hybridization techniques (for example, FISH) can be applied in concert to allow for direct empirical evaluation of the proposed conceptual framework. This approach links an important piece of the ecological literature, ecological stoichiometry, with the molecular front of the microbial revolution, in an attempt to provide new insight into how microbial physiology could constrain ecosystem processes.

Ecosystems↗
USGS2017

Prior knowledge-based approach for associating contaminants with biological effects: A case study in the St. Croix River basin, MN, WI, USA

Evaluating potential adverse effects of complex chemical mixtures in the environment is challenging. One way to address that challenge is through more integrated analysis of chemical monitoring and biological effects data. In the present study, water samples from five locations near two municipal wastewater treatment plants in the St. Croix River basin, on the border of MN and WI, USA, were analyzed for 127 organic contaminants. Known chemical-gene interactions were used to develop site-specific knowledge assembly models (KAMs) and formulate hypotheses concerning possible biological effects associated with chemicals detected in water samples from each location. Additionally, hepatic gene expression data were collected for fathead minnows ( Pimephales promelas ) exposed in situ , for 12 d, at each location. Expression data from oligonucleotide microarrays were analyzed to identify functional annotation terms enriched among the differentially-expressed probes. The general nature of many of the terms made hypothesis formulation on the basis of the transcriptome-level response alone difficult. However, integrated analysis of the transcriptome data in the context of the site-specific KAMs allowed for evaluation of the likelihood of specific chemicals contributing to observed biological responses. Thirteen chemicals (atrazine, carbamazepine, metformin, thiabendazole, diazepam, cholesterol, p-cresol, phenytoin, omeprazole, ethyromycin, 17β-estradiol, cimetidine, and estrone), for which there was statistically significant concordance between occurrence at a site and expected biological response as represented in the KAM, were identified. While not definitive, the approach provides a line of evidence for evaluating potential cause-effect relationships between components of a complex mixture of contaminants and biological effects data, which can inform subsequent monitoring and investigation.

Minnesota, Wisconsin↗
USGS2026

Weathering the ride: Associations of heat, smoke, precipitation, and ozone with bus ridership in Colorado

Weather and climate variability increasingly shape urban travel behavior, yet the short-term temporal dynamics and contextual modifiers of weather-transit relationships remain poorly understood. We analyzed system-wide hourly bus ridership in the Denver metropolitan area from June 2022 through September 2023 using fixed-effects negative binomial distributed lag nonlinear models with lags up to 24 h. Nonlinear exposure-response and lag-response functions were specified for hourly Universal Thermal Climate Index (UTCI) and precipitation, with additional indicators for daily ozone exceedance and wildfire smoke. Models adjust for hour-of-day, day-of-week, and month-year fixed effects, with standard errors clustered by day. Stratified analyses assess heterogeneity by time period, season, fare policy, and shelter availability, with interaction evaluated using Wald tests. Cold thermal stress was associated with the largest, most persistent reductions in ridership, with cumulative declines of −15.7% (95% CI: −26.0%, −4.0%) over 0-24 h. Precipitation was associated with sharp but transient reductions concentrated within 3-6 h (−8.4% to −9.8%), with little evidence of longer-term displacement. Heat associations were weaker and context-dependent, with modest short-run increases, but net same-day declines. Associations varied by time of day and season, were attenuated during the free-fare period, and differed by shelter availability. In contrast, ozone exceedance and wildfire smoke exhibited limited and inconsistent associations, with measurable reductions primarily during morning commute during heavier smoke conditions. Overall, transit ridership is more strongly associated with short-term weather exposure, than with ambient air quality. Fare policy and stop-level infrastructure modify these associations, highlighting actionable strategies to enhance transit resilience under increasing climate variability.

Colorado↗
USGS2009

A method to assess longitudinal riverine connectivity in tropical streams dominated by migratory biota

1. One way in which dams affect ecosystem function is by altering the distribution and abundance of aquatic species. 2. Previous studies indicate that migratory shrimps have significant effects on ecosystem processes in Puerto Rican streams, but are vulnerable to impediments to upstream or downstream passage, such as dams and associated water intakes where stream water is withdrawn for human water supplies. Ecological effects of dams and water withdrawals from streams depend on spatial context and temporal variability of flow in relation to the amount of water withdrawn. 3. This paper presents a conceptual model for estimating the probability that an individual shrimp is able to migrate from a stream's headwaters to the estuary as a larva, and then return to the headwaters as a juvenile, given a set of dams and water withdrawals in the stream network. The model is applied to flow and withdrawal data for a set of dams and water withdrawals in the Caribbean National Forest (CNF) in Puerto Rico. 4. The index of longitudinal riverine connectivity (ILRC), is used to classify 17 water intakes in streams draining the CNF as having low, moderate, or high connectivity in terms of shrimp migration in both directions. An in-depth comparison of two streams showed that the stream characterized by higher water withdrawal had low connectivity, even during wet periods. Severity of effects is illustrated by a drought year, where the most downstream intake caused 100% larval shrimp mortality 78% of the year. 5. The ranking system provided by the index can be used as a tool for conservation ecologists and water resource managers to evaluate the relative vulnerability of migratory biota in streams, across different scales (reach-network), to seasonally low flows and extended drought. This information can be used to help evaluate the environmental tradeoffs of future water withdrawals. ?? 2008 John Wiley & Sons, Ltd.

Aquatic Conservation: Marine and Freshwater Ecosys↗
USGS2022

The effect of scent lures on detection is not equitable among sympatric species

Context: Camera trapping is an effective tool for cost-efficient monitoring of species over large temporal and spatial scales and it is becoming an increasingly popular method for investigating wildlife communities and trophic interactions. However, camera trapping targeting rare and elusive species can be hampered by low detection rates, which can decrease the accuracy and precision of results from common analytical approaches (e.g., occupancy modeling, capture-recapture). Consequently, researchers often employ attractants to increase detection without accounting for how attractants influence detection of species among trophic levels. Aims: We aimed to evaluate the influences of a commonly used non-species-specific olfactory lure (i.e. sardines) and sampling design on detection of four species (i.e. bobcat [ Lynx rufus ], coyote [ Canis latrans ], raccoon [ Procyon lotor ], and eastern cottontail [ Sylvilagus floridanus ]) that represented a range of foraging guilds in an agricultural landscape. Methods: We set 180 camera stations, each for ∼28 days, during the summer of 2019. We set cameras with one of three lure treatments: (1) olfactory lure, (2) no olfactory lure, or (3) olfactory lure only during the latter half of the survey. We evaluated the influence of the lure at three temporal scales of detection (i.e. daily probability of detection, independent sequences per daily detection, and triggers per independent sequence). Key results: The lure tended to positively influence detection of coyotes and raccoons but negatively influenced detection of bobcats and eastern cottontails. The influence of the lure varied among temporal scales of detection. Conclusions: Scent lures can differentially influence detection of species within or among tropic levels, and the influence of a scent lure may vary among temporal scales. Implications: Our results demonstrate the importance of evaluating the influence of an attractant for each focal species when using camera data to conduct multi-species or community analyses, accounting for variation in sampling strategies across cameras, and identifying the appropriate species-specific temporal resolution for assessing variation in detection data. Furthermore, we highlight that care should be taken when using camera data as an index of relative abundance (e.g. as is commonly done with prey species) when there is variation in the use of lures across cameras.

Wildlife Research↗
USGS2025

Failure to meet the exchangeability assumption in Bayesian multispecies occupancy models: Implications for study design

Bayesian hierarchical models are ubiquitous in ecology. Random effect model structures are often employed that treat individual effects as deviations from larger population-level effects. In this way individuals are assumed to be "exchangeable" samples. Ecologists may address this exchangeability assumption intuitively, but might in certain modeling contexts ignore it altogether, including in situations where it may have large implications for study design. Multispecies occupancy models based on detection/non-detection data are an approach that can be utilized by those tasked with monitoring rare and endangered species because most literature suggests that, compared to single species occupancy models, improved parameter estimates are assured. Yet, we illustrate through a power analysis how sampling requirements to detect experimental treatment effects vary tremendously depending on whether the species exchangeability assumption is met. The degree to which species in a community respond similarly to covariates governs the ability to accurately estimate parameters using multispecies occupancy models. Detecting small or moderate changes in occupancy resulting from habitat restoration treatments may be impossible for small datasets (e.g., < 36 sampling locations, each surveyed < 8 times) even with a paired treatment-control design if the exchangeability assumption is violated. By contrast, when the assumption is met, small effects may be confidently estimated with as few as 12 sampling locations (6 pairs) and 6-8 survey events. Often, it may be impossible to know whether the exchangeability assumption is met. The statistical power needed to accurately estimate species-specific effects using detection/non-detection multispecies occupancy models depends on the unknown values of treatment effects and whether responses by species in the community diverge. When the species exchangeability assumption is violated, and at lower levels of sampling effort, multispecies occupancy models may provide worse inference than single species occupancy models.

BioRxiv↗
USGS2018

Landscape structure and temporal dynamic effects on Wintering Mallard abundance and distributions in the Mississippi alluvial valley

Context Management of wintering waterfowl in North America requires adaptability because constant landscape and environmental change challenges existing management strategies regarding waterfowl habitat use at large spatial scales. Migratory waterfowl including mallards (Anas platyrhynchos) use the lower Mississippi Alluvial Valley (MAV) for wintering habitat, making this an important area of emphasis for improving wetland conservation strategies, while enhancing the understanding of landscape-use patterns. Objectives We used aerial survey data collected in the Arkansas portion of the MAV (ARMAV) to explain the abundance and distribution of mallards in relation to variable landscape conditions. Methods We used two-stage, hierarchical spatio-temporal models with a random spatial effect to identify covariates related to changes in mallard abundance and distribution within and among years. Results We found distinct spatio-temporal patterns existed for mallard distributions across the ARMAV and these distributions are dependent on the surrounding landscape structure and changing environmental conditions. Models performing best indicated seasonal surface water extent, rice field, wetland and fallow (uncultivated) field abundance positively influenced mallard distribution. Rice fields, surface water and weather were found to influence mallard abundance. Additionally, are results suggest weather and changing surface water affects mallard presence and abundance throughout the winter, because the probability of mallard presence and abundance changed from the northern ARMAV in November to the southern ARMAV in January. Conclusions Using novel datasets to identify which environmental factors drive changes in regional wildlife distribution and abundance can improve management by providing managers additional information to manage land over landscapes spanning private and public lands. We suggest our analytical approach may be informative in other areas and for other wildlife species.

Arkansas↗
USGS2019

Optimal spatial allocation of control effort to manage invasives in the face of imperfect detection and misclassification

Imperfect detection and misclassification errors are often ignored in the context of invasive species management. Here we present an approach that combines spatially explicit models and an optimization technique to design optimal search and destroy strategies based on noisy monitoring observations. We focus on two invasive plants, melaleuca ( Melaleuca quinquenervia ) and Old World climbing fern ( Lygodium microphyllum ), which continue to cause important damages to the Everglades ecosystem. We present a methodological framework that combines Hidden Markov Random Field (HMRF, initially developed for image analysis) and linear programming to optimally search for invasive species. A benefit of this approach is that it accounts for the spatial structure of the system by using a spatially explicit modeling approach (i.e. HMRF), and does not require repeated visits to model the probability of occurrence of species. We found on simulated cases that our approach can lead to substantial improvements in control efficiency when compared to state of the art model-free approaches. For example, in the case of the old world fern, simulations showed that the optimal strategy would allow managers to control up to 34% more sites than with model-free approaches that ignored misclassification and imperfect detection. For melaleuca it was possible to control up to 20% more sites. The vast increase in imagery data obtained from different sources (e.g. unmanned aerial systems, and satellite) provides great opportunities to improve management of natural resources by applying modern computational methods such as the one we present. Our approach can substantially increases the efficiency of invasive species control by accounting for imperfect detection, misclassification error and the spatial structure of the system. Our approach is applicable to other systems and problems, for example it could be applied to the control of plant pathogens, or optimal extraction of resources (e.g. minerals or biological resources).

Ecological Modelling↗
USGS2022

A review of supervised learning methods for classifying animal behavioural states from environmental features

Accurately predicting behavioural modes of animals in response to environmental features is important for ecology and conservation. Supervised learning (SL) methods are increasingly common in animal movement ecology for classifying behavioural modes. However, few examples exist of applying SL to classify polytomous animal behaviour from environmental features especially in the context of millions of animal observations. We review SL methods (weighted k -nearest neighbours; neural nets; random forests; and boosted classification trees with XGBoost) for classifying polytomous animal behaviour from environmental predictors. We also describe tuning parameter selection and assessment strategies, approaches for visualizing relationships between predictors and class outputs, and computational considerations. We demonstrate these methods by predicting three categories of risk to bald eagles from colliding with wind turbines using, as predictors, 12 environmental state features associated with 1.7 million GPS telemetry data points from 57 eagles. Of the SL methods we considered, XGBoost yielded the most accurate model with 86.2% classification accuracy and pairwise-averaged area under the ROC curve of 90.6. Computational time of XGBoost scaled better to large data than any other SL method. We also show how SHAP values integrated in the R package ( xgboost ) facilitate investigation of variable relationships and importance. For big data applications, XGBoost appears to provide superior classification accuracy and computational efficiency. Our results suggest XGBoost should be considered as an early modelling option in situations where the intent is to classify millions of animal behaviour observations from environmental predictors and to understand relationships between those predictors and movement behaviours. We also offer a tutorial to assist researchers in implementing this method.

Methods in Ecology and Evolution↗
USGS2023

Translating stakeholder narratives for participatory modeling in landscape ecology

Context Engaging stakeholders in research is needed for many of the sustainability challenges that landscape ecologists address. Involving stakeholders’ perspectives through narratives in participatory modeling fosters better understanding of the problem and evaluation of the acceptability of tradeoffs and creates buy-in for management actions. However, stakeholder-driven inputs often take the form of complex qualitative descriptions, rather than model-ready numerical or categorical inputs. Objectives Translating narratives into models, model parameters, or scenarios is essential for leveraging stakeholder knowledge and engagement. Drawing from varied experiences to identify lessons learned and pitfalls, we address the practice of translating narratives into models and using those narratives to interpret and communicate results. Methods We drew from seven participatory landscape ecology projects across North America to synthesize lessons for the inclusion of stakeholder narratives in modeling studies. Results We offer 8 lessons as practical guidance for other landscape ecologists to move the science beyond a unilateral focus on ecological systems and to maximize the benefits of landscape sustainability science. Conclusions These lessons are starting points, as real projects are complex, nuanced, and sometimes contradictory. Translating narratives into models is important for addressing complex sustainability challenges; we hope that these starting points are helpful to those foraying into this type of research.

Landscape Ecology↗