ScholarWorks

ScholarWorks is an open access repository for the capture of the intellectual work of Montana State University (MSU) in support of its teaching, research and service missions. MSU ScholarWorks is a central point of discovery for accessing, collecting, sharing, preserving, and distributing knowledge to the Montana State University community and the world.

MSU PHOTOS

Communities in ScholarWorks

Select a community to browse its collections.

Now showing 1 - 1 of 1

Scholarship & Research

Recent Submissions

Leveraging an observed-data likelihood improves the use of machine learning labels in a Bayesian hierarchical model for bioacoustic data
(Institute of Mathematical Statistics, 2025-12)
Oram, Jacob K.
;
Banner, Katharine M.
;
Stratton, Christian
;
Hoegh, Andrew
;
Irvine, Kathryn M.
Classification of massive datasets by machine learning (ML) algorithms is promising for many scientific domains, especially wildlife monitoring programs that rely on passive acoustic surveys for detecting species. However, treating ML-predicted class labels (e.g., species identity) as truth biases inferences of focal parameters within common modeling frameworks. One solution is to model the misclassification process explicitly using human-validated true-class labels for a subset of observations. Validation by experts can present a substantial bottleneck in otherwise efficient workflows that use ML predictions. Bioacoustics practitioners seek guidance on both the quantity and process for selecting ML-labeled data to validate by an expert. We derive an alternative model formulation that jointly models human-validated and ML-predicted class labels with an observed-data likelihood (ODL) and use empirically informed simulations motivated by a real-data application to explore different probability designs for selecting class labels for validation. Simulation results suggest that with smaller validation sets the ODL formulation increases computational speed and reduces estimation error compared to a default MCMC data augmentation routine. Our methodology is transferable to applications that treat predictions from classification algorithms as the response variable of interest.
Two classes of amine/glutamate multi-transmitter neurons innervate Drosophila internal male reproductive organs
(eLife Sciences Publications, 2025-10)
Chaverra, Martha
;
Toney, John Paul
;
Dardenne-Ankringa, Lizetta D.
;
Tolleson Knee, Jace
;
Morris, Ann R.
The essential outcome of a successful mating is the transfer of genetic material from males to females in sexually reproducing animals from insects to mammals. In males, this culminates in ejaculation, a precisely timed sequence of organ contractions driven by the concerted activity of interneurons, sensory neurons, and motor neurons. Although central command circuits that trigger copulation have been mapped, the motor architecture and the chemical logic that couple specific neuronal subclasses to organ-specific contractility, seminal fluid secretion, and sperm emission remain largely uncharted. This gap in knowledge limits our ability to explain how neural circuits adapt to varying contexts and how their failure contributes to infertility. Here, we present an in-depth anatomical and functional analysis of the motor neurons that innervate the internal male reproductive tract of Drosophila melanogaster. We identify two classes of multi-transmitter motor neurons based on neurotransmitter usage, namely octopamine and glutamate neurons (OGNs) and serotonin and glutamate neurons (SGNs), each with a biased pattern of innervation: SGNs predominate in the accessory glands, OGNs in the ejaculatory duct, with equal contributions of each to the seminal vesicles. Both classes co-express vesicular transporters for glutamate (vGlut) and amines (vMAT), confirming their dual chemical identity. Their target organs differentially express receptors for glutamate, octopamine, and serotonin, suggesting combinatorial neuromodulation of contractility. Functional manipulations show that SGNs are essential for male fertility but OGNs are dispensable. Glutamatergic transmission from both classes is also dispensable for fertility. These findings provide the first high-resolution map linking multi-transmitter motor neurons to specific reproductive organs, reveal an unexpected division of labor between serotonergic and octopaminergic signaling pathways, and establish a framework for dissecting conserved neural principles that govern ejaculation and male fertility.
Building an inclusive conservation vision founded upon ecological values and social opportunities
(IOP Publishing, 2025-08)
McKinley, Peter S.
;
Cheek, Lia
;
Belote, R. Travis
We developed a spatially explicit model for the eastern United States to help identify where we work to conserve a network of ecologically important lands with the input of communities often excluded from conservation planning. We built multiple individual and composite maps from ecological data selected from four broad themes: ecological integrity, connectivity, biodiversity, and ecosystem services. Datasets selected from these themes identify lands important for a conservation network resilient to global change stressors that provide important functions upon which people depend. We also built multiple individual and composite maps using spatial data from existing efforts to measure social conditions constructed around three broad themes: frontline communities, historically marginalized populations, and people who experience the impacts of climate change most accurately. We view these areas as having high social opportunity through improvement of the environmental conditions experienced by these communities. We also hope to engage a greater representation of values, experience, and knowledge held by communities not typically part of conservation planning. We assert that these aims can only be achieved by amplifying excluded community voice and leadership when developing approaches to conservation. This spatially explicit social-ecological model is comparable to models we have built to facilitate development of various collaboratives and initiatives in our place-based work. This is a type of decision support tool, not a decision maker. We present a broad spatial summary of three conditions: 1) co-occurring high social opportunity and nationally significant ecological value, 2) areas of high social opportunity, and 3) areas of high ecological value. These analyses are presented as a foundation for large scale and collaborative conservation planning that seeks to conserve key ecological areas while addressing the needs of a broader spectrum of people. We envision regional conservation efforts that support collective ecological and social well-being. Our framework and data can also be rescaled to smaller extents to identify projects where social and ecological well-being might intersect in local areas. While our combined spatial data synthesizes myriad information, our collection of the individual criteria can serve as a geospatial library of resources available to regional and local conservation efforts. While we intend for this work to guide conservation strategies including land protection, land management, and land stewardship initiatives in conjunction with social initiatives, this work is a tool and not itself a method or blueprint for the challenging work ahead.
Trait Mapping Utilizing a Newly Constructed Genome for Allohexaploid Invasive Eurasian Watermilfoil (Myriophyllum spicatum) Reveals a Non-Target Site QTL Associated With Fluridone Resistance
(Wiley, 2026-01)
Hannay, Del
;
Chorak, Gregory M.
;
Harkess, Alex
;
Clevenger, Josh
;
Cuperus, Josh T.
Herbicides are a valuable tool in agricultural ecosystems to manage nuisance species. Due to the reliance on herbicides for weed control, herbicide resistance is a growing concern. Herbicides are also used extensively in aquatic and natural systems, but the genetics and evolutionary dynamics of resistance are not as frequently incorporated into management plans in these systems. In Eurasian watermilfoil, a widespread and heavily managed invasive aquatic weed in the United States, clonal lineages have been characterized as resistant to fluridone, a commonly used phytoene desaturase (PDS)-inhibitor herbicide. In order to locate genomic loci associated with herbicide resistance, we created an F2 mapping population segregating for fluridone resistance. Using this population, we examined the pds gene for amino acid alterations in resistant individuals and performed bulk segregant analysis between the highly resistant and susceptible F2 individuals. Additionally, we compared pds gene expression between resistant and susceptible strains in control and treated environments using RT-qPCR. We found no evidence of amino acid alterations to the pds gene in fluridone resistant individuals or increased pds expression in the resistant strain, either in the presence or absence of fluridone. Our QTL mapping identified a putative QTL on chromosome seven, while the gene encoding fluridone's target molecule, phytoene desaturase (PDS) is located on chromosomes 10–12. Our results indicate that fluridone resistance in the Eurasian watermilfoil strain isolated from Lake Lansing, MI, is due to at least one non-target site mechanism. Characterizing mechanisms of herbicide resistance within invasive plants enables effective and thoughtful herbicide usage, as well as the development of diagnostic biomarkers for resistance in unknown populations.
A Comparison of MAVEN SIR Observations With the Stationary WSA-ENLIL Solar Wind Model
(Wiley, 2025-12)
Henderson, Sarah
;
Filwett, Rachael
;
Owen, Sophia
;
Halekas, Jasper
;
Gruesbeck, Jacob
Predicting times of arrival and properties of space weather events, such as coronal mass ejections and stream interaction regions (SIRs), has become an important focus of the space physics community within recent years. Extensive efforts have been undertaken to model these space weather events throughout the heliosphere in order to better understand their properties and predict how and when they will impact planetary environments. In this study, we compare in situ solar wind parameters during SIRs measured by the Mars Atmosphere and Volatile EvolutioN (MAVEN) spacecraft to parameters generated by the Wang-Sheeley-Arge (WSA)-ENLIL stationary solar wind model. We compare times of arrival, end times, event duration, solar wind parameters, and magnetic compression ratios as measured by MAVEN versus WSA-ENLIL using a previously compiled 9-year catalog of SIR observations. We find that WSA-ENLIL, on average, predicts earlier times of arrival and longer duration SIRs than what is observed by MAVEN. We examine how in situ solar wind parameters compare to those predicted by WSA-ENLIL and find that solar wind proton density and magnetic field magnitude are slightly underpredicted by the model, while solar wind speed is well predicted. We also find that the magnetic compression ratio predicted by WSA-ENLIL is higher than MAVEN by an average of ∼ 26%. We examine the effects of planetary geometry on the modeling outputs and find that certain parameters are more adversely impacted than others depending on the alignment of Earth and Mars.