Justin Klone, an analyst and graduate student at the University of Missouri research project, developed a method to track the spread of invasive Karelly Pear trees in central Missouri using freely available satellite imaging and machine learning techniques. Krohn initially recorded the exact location of Callery Pear Trees in Columbia, Missouri using a GPS device, and then used this data to train a machine learning model that could distinguish trees from surrounding vegetation based on variations in light reflectance. This study revealed a higher prevalence of callery pear trees in suburban areas characterized by more open land, in contrast to the limited spread observed in more populated urban areas in Colombia. This study shows that these trees prefer disturbed habitats, particularly those adjacent to new residential developments and roads, suggesting a correlation between artificial land use and invasive species establishment. This methodology provides a low-cost alternative to traditional remote sensing technologies using drones or aircraft images to monitor the distribution of invasive species and predict future expansion within Missouri ecosystems.
Invasive species tracking
The University of Missouri research initiative, led by research project analyst and graduate student Justin Krone, is pioneering a cost-effective methodology to track the surge in invasive Karely pear trees in central Missouri. This project addresses important ecological concerns Pyrus Calleryana – commonly known as Karelinas – exhibits rapid growth, poses a threat to native flora through competitive exclusion, exhibits structural debilitating and increases the likelihood of damage caused by storms. In response to ecological impacts, Missouri legislative measures reflecting that of other states, require a robust monitoring strategy to enact a ban on the sale of these trees, assess the effectiveness of these measures, and inform future conservation efforts. The core of Krohn's approach lies in the integration of freely accessible satellite imagery and machine learning algorithms. This is a subset of artificial intelligence that focuses on allowing you to learn from data without explicit programming. Traditional methods of invasive species mapping rely on aircraft investigations carried out via drones or aircraft, resulting in significant financial costs. Krohn's approach avoids these limitations by leveraging publicly available data sources and automated analytics. The methodology involves establishing a ground truth data set first. The exact location of the pear tree in Colery, Columbia, Missouri, is recorded using GPS devices and provides labeled data for training machine learning models. The subsequent phase involved training the algorithm to distinguish the karelinis tree from its surrounding environment, based on the signature of the spectral. This is the variation in the reflection of electromagnetic radiation across different wavelengths. Vegetation exhibits unique spectral properties that depend on factors such as chlorophyll content, leaf structure, and moisture. Machine learning models learn to identify these nuances and allow automatic detection of karelinis trees in satellite images. This process known as monitored learning requires a significant amount of labeled data to achieve high accuracy, highlighting the importance of the initial ground truth exercise. Analysis of the collected data revealed discernable patterns in the distribution of invasive cultivated pear trees within Columbia, Missouri. A higher prevalence was observed in suburban areas characterized by more open land, in contrast to dense, densely populated urban centres where opportunities for spread are constrained. This finding is consistent with ecological principles such that disturbed habitats, such as those associated with new housing developments and roadside edges, provide ideal conditions for the establishment and proliferation of invasive species. This study suggests that continuing urban expansion could exacerbate the spread of Colely pears, highlighting the need for a positive management strategy. The findings have been published at relevant scientific conferences and published in peer-reviewed journals, and are intended to contribute to a broader knowledge of invasive species ecology and management. Funding for the project was provided through a research grant from the University of Missouri. ##Machine Learning Applications
The application of machine learning to detect and monitor invasively developed pear trees represents a significant methodological advancement over traditional ecological research techniques. Justin Krone, a research project analyst and graduate student at the University of Missouri, led the development of a supervised learning algorithm trained on high-resolution satellite imagery. This approach avoids the logistics and financial burdens associated with extensive field surveys or aerial image acquisition using drones or aircraft, providing a scalable and cost-effective solution for large-scale area monitoring. The core of the methodology relies on the principles of spectral analysis. Different plant species exhibit unique reflection patterns throughout the electromagnetic spectrum determined by their biochemical composition and structural properties. The monitored learning process began with a meticulous movement of terrestrial truths in which the exact location of the Karelinis tree in Columbia, Missouri was recorded using GPS devices and created a spatially referenced dataset. This labeled data serves as a training set for machine learning models, and in particular can correlate spectral signatures (variability in light reflections) with the presence of invasive species. This algorithm employs leverage techniques such as random forests and support vector machines to process high-dimensional data and identify complex patterns within spectral information. These algorithms are particularly suitable for image classification tasks. The goal is to assign each pixel to a specific class. The success of machine learning models depends on their ability to generalize from training data, and accurately identifying the karelinis tree in invisible images. This requires careful consideration of factors such as image resolution, atmospheric conditions, and spectral characteristics of other vegetation types present in the landscape. Krohn's Research meticulously addresses these challenges through data preprocessing techniques and algorithmic parameter tuning, ensuring a robust and reliable detection rate. The resulting model provides a spatially explicit map of the distribution of karelinas, facilitating targeted management interventions, and allows predictive modeling of future spreads. The findings published at the Science Conference are intended to contribute to a broader understanding of invasive species ecology and management strategies and to be published in peer-reviewed journals. Funding for this innovative research was provided through a research grant from the University of Missouri, demonstrating institutional support for technically advanced ecological surveillance. ##Suburban prevalence
A study conducted by Justin Krone, a research project analyst and graduate student at the University of Missouri, reveals a statistically significant correlation between callery pear tree prevalence within Columbia, a metropolitan area of Missouri and suburban land use patterns. Analysis of spatially referenced data collected via GPS logging of individual trees shows a significantly higher density of invasive cultivated pear trees in suburban landscapes compared to more densely developed urban centers. This distribution is attributed to the availability of disturbed habitats characteristic of suburban expansion, particularly those adjacent to new housing developments, road edges, and areas adjacent to previous farmlands. These fragmented landscapes provide ideal conditions for the establishment and rapid growth of invasive species, exploiting soil disorders and increasing light availability. The observed suburban prevalence is further explained by the ecological properties of invasive cultivated pears. This species exhibits high tolerance to a variety of soil conditions, exhibits rapid growth rates, allowing rapid colonization of disturbed areas. Furthermore, the dispersal mechanism of callery pears, primarily through bird seed dispersal and nutrient spread, is particularly effective in suburban matrices where fragmented habitats promote long-distance seed transport and the establishment of new populations. This study highlights that the open nature of suburban landscapes, coupled with frequent obstacles, creates a positive feedback loop and encourages the continued spread of invasive species. Krohn's study adopted freely available satellite images processed using machine learning algorithms to extrapolate these findings beyond locations logged in the initial GPS. The resulting spatially explicit map of the distribution of karelinas provides valuable tools for land managers and conservation practitioners. Identifying areas with high concentrations of invasive species can more efficiently implement target management interventions such as selective removal and precautions. This approach contrasts with the extensive traditional control efforts, often ineffective and costly. The methodology developed by Krohn provides a scalable and cost-effective solution for monitoring and managing invasive species in larger geographical areas. The implications of this study go beyond the immediate context of Columbia, Missouri. The observed pattern of suburban prevalence may apply to other metropolitan areas in the Midwest and the Eastern US, where invasive adoptive pear trees are prevalent. Understanding the relationship between land use patterns and invasive species spread is important for developing effective conservation strategies in rapidly urbanising landscapes. This study highlights the need for a proactive management approach to address the root causes of invasion, such as habitat disruption and fragmentation, rather than simply responding to established invasions. Further research is needed to investigate the long-term ecological consequences of invasive callery pear domination in suburban ecosystems and to assess the effectiveness of various management interventions.
