River ecosystems and their plant communities are a critical component to the landscape of Mississippi and invasion represents a substantial threat to these systems. This study focused on the Pascagoula River, the Pearl River, and the Tennessee-Tombigbee Waterway, whose hydrology range from largely unaltered to highly altered. The rivers’ aquatic and riverine communities were measured using point surveys and site community compositions were assessed using nonmetric multidimensional scaling (NMDS). We found that sites on the Pascagoula and Pearl rivers were similar to each other while being dissimilar to the TTW sites, and that coastal systems on the Pascagoula and Pearl Rivers separated from the noncoastal sites. The effects of latitude and river system on species sample richness (s) were assessed using a linear mixed-effects model. The effects of latitude and native species richness (sn) on the presence/absence of introduced species (pai) were assessed using a generalized linear model. Results showed that across all river systems, decreasing latitude increased s. These results suggest that although these rivers have differences in their community composition, the effect of latitude on s is strong enough to be exhibited consistently in all three rivers. Additionally, we found that sn (positive) and latitude (negative) had substantial effects on pai with sn being the stronger predictor. These findings are consistent with the well-supported “rich get richer” hypothesis that posits a positive relationship between native plant richness and introduced plant richness. This pattern describes the broad and indirect association between native species richness and susceptibility to invasion.

Mosquito fern (Azolla caroliniana) is a free-floating aquatic fern capable of covering water bodies and outcompeting submersed macrophytes, potentially leading to a loss of biodiversity. Limited evidence suggests that mosquito fern can be controlled with the contact herbicide diquat; however, peer-reviewed literature regarding effects of other contact herbicides labeled for use in aquatic environments on mosquito fern is lacking. The purpose of this work was to conduct two trials to determine the effects of foliar (trial 1) and submersed (trial 2) applications of contact herbicides on mosquito fern. In trial 1, foliar applications of the contact herbicides flumioxazin (0.42 and 0.21 kg ai ha(-1)), carfentrazone-ethyl (0.21 and 0.11 kg ai ha(-1)), endothall (2.39 and 1.20 kg ai ha(-1)), diquat (4.52 and 2.26 kg ai ha(-1)), and copper (1.47 and 0.74 kg ai ha(-1)) were administered and biomass assessed 8 wk after treatment (WAT). In trial 2, submersed applications of flumioxazin (0.4 and 0.2 mg ai L-1), carfentrazone-ethyl (0.2 and 0.1 mg ai L-1), endothall (5.0 and 2.5 mg ai L-1), diquat (0.37 and 0.19 mg ai L-1), and copper (1.0 and 0.5 mg ai L-1) were administered and assessed 8 WAT. Foliar treatments were applied at a target diluent rate of 935.4 L ha(-1); all foliar herbicide treatments included a 1% v:v nonionic surfactant. At 8 WAT, all foliar treatments reduced mosquito-fern biomass compared with nontreated plants, but only high rates of flumioxazin, carfentrazone-ethyl, and both diquat rates reduced biomass 100%. All submersed herbicide treatments except copper reduced mosquito-fern biomass by 8 WAT, but diquat was the only treatment to provide 100% biomass reduction. To our knowledge, this is the only work to document mosquito fern biomass reduction by the herbicides flumioxazin, copper, carfentrazone-ethyl, or endothall. This work should be validated on field populations of mosquito fern before recommendation for operational use.

Understanding the specific nutrient and biomass allocation strategies of wetland plants is crucial for selecting suitable species or combinations of species for wetland restoration or creation. To investigate differences in growth and nutrient uptake, we measured the above- and belowground biomass of 520 individuals from four species grown in single and paired mesocosms over three growing seasons. Key growth parameters—including biomass, maximum height, number of leaves, and culms—were recorded, along with tissue nutrient content (C, N, P, K, Mg, Ca, S, B, Cu, Fe, Mn, Zn) and analyzed using inductively Coupled Plasma Optical Emission Spectrometry (ICP-OES). Juncus effusus in single-species cultures produced a higher average aboveground biomass and culm production than in paired cultures, where this species increased its own Ca, Mg, Cu, and Mn uptake. Species paired with J. effusus exhibited higher total and aboveground biomass, maximum height, and culm production (Typha latifolia); aboveground biomass and culm production (Phragmites australis); and culm production (Schoenoplectus tabernaemontani) than they did when grown alone. Additionally, S. tabernaemontani paired with J. effusus demonstrated higher nutrient uptake, while P. australis in paired culture enhanced its carbon concentration in aboveground tissues. We also observed strong correlations among nutrient concentrations, particularly in the case of P. australis. These findings highlight the importance of species selection in wetland restoration and creation, as specific species interactions, especially involving J. effusus, can enhance biomass production and nutrient uptake. This research offers valuable insights for optimizing plant pairings to improve nutrient mitigation, which could support the development of advanced ecological modeling for wetlands.

Both aquatic flora and Mississippi flora are underrepresented in herbarium collections. This paper reports nine noteworthy collections of 6 aquatic and wetland taxa in Alabama, Arkansas, and Mississippi. These taxa include 5 angiosperms and 1 charophyte. Two of these taxa are introduced to the United States, while the rest are presumed native. Lychnothamnus barbatus, Utricularia tenuicaulis, and Vallisneria x pseudorosulata are new to the flora of Mississippi and Cyperus blepharoleptos is new to the flora of Arkansas.

ABSTRACT Within the study of aquatic invasive species, small aquatic ecosystems are often neglected, despite representing most global freshwater bodies. This study uses community composition and environmental and geographic factors to explain the occurrence of invasive species in small lakes in the southeastern United States. Four invasive species widespread in the southeastern United States were selected as the focus of this study: Alternanthera philoxeroides, Cyperus blepharoleptos, Panicum repens, and Triadica sebifera. The aquatic plant communities of the lakes were surveyed using littoral zone point sampling. Generalized linear models for each species were fit with the probability of occurrence (Pocc) as the response variable and Secchi depth, plant species diversity (α‐diversity), point richness, perimeter, latitude, and longitude as potential predictors; all predictors were subjected to model selection to define the best‐fit models. All best‐fit models were strongly predictive with area under the receiver operating characteristic curve values > 0.80. Plant species diversity was positively correlated with Pocc of A. philoxeroides, P. repens, and T. sebifera. Latitude was negatively correlated with Pocc of P. repens and T. sebifera. Perimeter was negatively related to Pocc of A. philoxeroides. Secchi depth was negatively related to the Pocc of C. blepharoleptos. Although plant species diversity and latitude were most commonly predictive, Pocc was usually explained by multiple predictors, suggesting that these relationships are best explained with multiple environmental factors.

In this research, a cost-effective NVIDIA Jetson based computer vision system has been developed to detect and classify eight aquatic invasive plants commonly found in the southeastern United States. A comprehensive image dataset comprising 1,963 visible spectrum high-resolution images was collected, representing eight plant classes: 1) Alligator weed, 2) Cuban bulrush, 3) Giant salvinia, 4) Primrose, 5) Torpedo grass, 6) Water hyacinth, 7) Water lettuce, and 8) Water lily. The following six deep learning models were trained and evaluated to detect and classify plant classes: MobileNetV2, ResNet50, InceptionV3, EfficientNet, VGG19, and ViT. These trained models were then transferred to a NVIDIA Jetson Nano microcomputer interfaced with an Arducam IMX 219 visible camera. The computer vision hardware was extensively tested in the field. We observed that there is a trade-off between classification accuracy and inference time. ResNet50 and InceptionV3 achieved the highest accuracy in real-world testing. A notable reduction in performance was seen in models such as MobileNetV2 and EfficientNet. This research demonstrates the potential of deep learning models for automating the monitoring of invasive species, offering a cost-effective, scalable solution for environmental monitoring and the management of aquatic ecosystems.

Alternanthera philoxeroides (Mart.) Griseb. (Amaranthaceae: Caryophyllales) is an aquatic invasive weed from South America with a long history of biological control. The well-studied Agasicles hygrophila Selman Vogt, 1971 (Coleoptera: Chrysomelidae) successfully controls A. philoxeroides in some parts of its invaded range, but is unsuitable in other areas due to its intolerance to cold temperatures. Amynothrips andersoni O’Neill, 1968 (Thysanoptera: Phlaeothripidae) has shown greater tolerance to cold temperatures, but no research has been conducted to determine its ecological niche with respect to A. philoxeroides. The aim of this study is to predict the environmental niches of A. andersoni and A. hygrophila and their overlap with that of A. philoxeroides in the North and South America under current and future climate scenarios. Accordingly, niche models were constructed in MaxEnt for all three species using environmental variables from the current climate and under two future climate scenarios (SSP1-2.6 and SSP5-8.5) for the year 2040. The niche overlap between the two biological control agents and the host were estimated for all three scenarios. Under both future climate scenarios, the total niche of A. philoxeroides is predicted to decrease by up to 10

Cyanobacterial blooms pose significant threats to aquatic ecosystems and public health due to their ability to release harmful toxins, degrade water quality, disrupt aquatic habitats, and endanger human and animal health through contact or consumption of contaminated water. Monitoring phycocyanin (PC), a pigment unique to cyanobacteria, offers a reliable method for detecting and quantifying these blooms, enabling timely interventions to mitigate their impacts. This study aimed to evaluate ten machine learning algorithms (MLAs) for assessing the spatiotemporal variations of cyanobacterial concentrations over an oyster reef in the Western Mississippi Sound (WMS) using remotely sensed imagery from uncrewed aircraft systems (UAS) and in-situ PC concentrations measured by an autonomous surface vessel (ASV). The study further investigated the influence of river discharge and climatic variables on cyanobacterial concentrations using a time-series of cyanobacteria maps. To derive the most accurate PC retrieval model, a comprehensive set of 85 features was initially generated, including individual spectral bands, band ratios, multiple vegetation indices, and three-band indices. Feature selection was performed using a two-step approach that combined Sequential Backward Floating Selection (SBFS) and Exhaustive Feature Selection (EFS). SBFS was first used to iteratively remove features and optimize model performance, while EFS evaluated all possible combinations of the features identified by SBFS to select the best subset. Among the ten MLAs tested, Extreme Gradient Boosting emerged as the top-performing model, achieving an R2 of 0.835, a root mean square deviation of 0.419 μg/l, an unbiased mean absolute relative difference of 0.176 μg/l, and an average percentage difference of 18.072 % in retrieving PC concentration. The novelty of this study lies in its data-driven approach to identifying the most suitable machine learning algorithm and feature subsets for PC retrieval, thereby enhancing the accuracy and robustness of the developed algorithm. The time-series analysis revealed substantial variations in cyanobacterial concentration in the WMS from 2018 to 2022. The highest average concentration occurred in 2019, coinciding with the introduction of diverted Mississippi River water through the Bonnet Carré Spillway, which triggered an unprecedented cyanobacterial bloom. Furthermore, the average PC concentration was consistently higher during the summer months, likely due to elevated air temperatures and increased sunlight promoting cyanobacterial growth. The methodology developed in this study improves the quantitative monitoring of cyanobacterial blooms using UAS imagery and provides valuable insights for future water quality monitoring initiatives in other regions.

Chlorophyll-a (Chl-a) is a critical biological indicator of the eutrophic state of water bodies, emphasizing the importance of its detailed characterization and continuous monitoring. This study evaluated the performance of 10 widely used Machine Learning (ML) algorithms in deriving the spatiotemporal distribution of Chl-a from uncrewed aircraft systems (UAS) imagery. Field data for Chl-a algorithm development were collected simultaneously using an Autonomous Surface Vessel (ASV), a hand-held radiometer, and a UAS in the Western Mississippi Sound (WMS). To ensure the algorithms were developed using accurate remote sensing reflectance data, the spectral response function of the UAS was applied to the radiometer measurements. An initial dataset comprising of 85 variables was compiled, including individual spectral bands, band ratios, vegetation indices, and three-band indices. Two feature selection techniques-Sequential Backward Floating Selection and Exhaustive Feature Selection-were employed to identify the optimal subset of variables. These techniques reduced the original d-dimensional feature space to a k-dimensional space by iteratively evaluating all possible feature combinations and selecting those that achieved the highest R2 scores for each ML algorithm. Model performance was further validated using a separate dataset and assessed through metrics such as Root-MeanSquare Difference (RMSD), Mean Absolute Relative Difference (MARD), and Average Percentage Difference (APD). Among the 10 ML algorithms tested, the extreme gradient boosting (XGB) algorithm demonstrated superior performance, achieving the highest R2 score of 0.848. Using a combination of two band ratios, three vegetation indices, and three three-band indices, the XGB algorithm achieved RMSD, MARD, and APD values of 0.538 mu g/L, 0.407 mu g/L, and 9.83 %, respectively. The XGB algorithm was subsequently applied to a time series of UAS imagery to generate Chl-a concentration maps, which revealed the spatiotemporal distribution of Chl-a across the study area. The methodology developed in this study provides a robust framework for monitoring Chla in the coastal waters of the WMS using UAS imagery. Furthermore, the techniques and findings provide valuable insights for advancing water quality monitoring efforts in other regions.

Vallisneria x pseudorosulata S. Fujii & M. Maki is an invasive aquatic weed that has recently become a major issue within the U.S. Southeast. Vallisneria x pseudorosulata is a hybrid between two nonnative eelgrass species (Vallisneria spiralis L. and Vallisneria denseserrulata Makino) and has rapidly overtaken water bodies in Tennessee, Alabama, and Florida. This hybrid can reproduce rapidly through offshoot formation and floating propagules capable of drifting large distances before establishing. Vallisneria x pseudorosulata has been previously found in Japan and is thought to have been introduced in the United States by the aquarium trade or through dumping.

Cuban bulrush (Oxycaryum cubense [Poepp. & Kunth] Lye) is an invasive floating aquatic plant that causes negative ecological and economic impacts in the southeastern United States. Temperatures in the United States have increased over recent decades which can result in geographic expansion of invasive plants in North America. Accumulated degree-days (ADD) were utilized to develop predictive models (state and regional models) for Cuban bulrush growth from harvested biomass collected over one year in Mississippi, Louisiana, and Florida. Peak emergent biomass occurred from early to mid-fall (September-October) with growth continuing into winter. Accumulated degree days needed for Cuban bulrush to reach peak emergent biomass ranged from 6,469 (Mississippi), 7111 (regional), 7,643 (Florida), and 7,903 (Louisiana). Calendar days needed for Cuban bulrush to reach peak emergent biomass ranged from 292 (Mississippi) to 334 (Florida). Base temperature thresholds for Cuban bulrush were -6 C, -3 C, -3 C, and -2 C for Mississippi, Louisiana, regional, and Florida models respectively. The models suggest Cuban bulrush has a tolerance to lower air temperatures that could allow for survival in moderate winter conditions. Overall, model predictability was less accurate for populations further south (Florida) due to warmer winter temperatures, year-round growth, and difficulty defining when peak emergent biomass occurred. Results from this study indicate that Cuban bulrush growth is greater in warmer temperatures, though low base temperature thresholds suggest this species may be capable of expanding its invaded range to cooler climates beyond the southeastern United States.

Alligatorweed [Alternanthera philoxeroides (Mart.) Griseb.; Amaranthaceae] is a globally problematic, aquatic invasive weed with a long history as a target for control efforts. Although chemical and biological control methods have been widely studied to manage alligatorweed infestations, many research questions remain unanswered. This paper seeks to assess the efficacy of two understudied alligatorweed control methods: submersed herbicide applications and biological control with alligatorweed thrips (Amynothrips andersoni O’Neill 1968; Thysanoptera: Phlaeothripidae). These assessments were carried out in mesocosm experiments, in two stages. The first stage tested five herbicides applied as submersed injections at two different rates, and the second tested the same five herbicides alone and in combination with alligatorweed thrips biological control. The submersed herbicides used in this study were penoxsulam, bispyribac-sodium, imazamox, fluridone, and topramezone. The control effect of these treatments was measured as percent biomass reduction 12 weeks after treatment. These data showed that, with the exception of bispyribac-sodium, submersed herbicide application was generally successful at reducing alligatorweed biomass. Also, thrips biological control was broadly effective at reducing alligatorweed biomass. However, these data did not identify a specific herbicide whose control was significantly benefitted by thrips biological control at the rates these herbicides were applied. While the results of this study show promise for combining submersed herbicides and alligatorweed thrips for integrated alligatorweed management, questions remain regarding this combined control strategy including whether or not these results translate to the field.

Cuban bulrush [ Oxycaryum cubense (Poepp. & Kunth) Lye] and water hyacinth [ Eichhornia crassipes (Mart.) Solms] cause major ecological and economical impacts in the southeastern United States. These species are commonly associated with each other because of the epiphytic nature of Cuban bulrush, as it utilizes water hyacinth (and other floating plants or objects) as a colonization substrate. Increasing global temperatures may allow for the northward expansion of both species in North America. The purpose of this study was to model plant growth (i.e., biomass) as a function of accumulated degree-days (ADD) to predict peak biomass of both species growing in Lake Columbus, Mississippi. Models suggested water hyacinth and Cuban bulrush life cycles were asynchronous, thus ADD calculations did not occur over the same set of dates. Water hyacinth biomass peaked in September (estimated 5,399 ADD), and Cuban bulrush biomass peaked in February (6,549 ADD). Estimated base temperature threshold at which water hyacinth growth occurs is - 1 C, while Cuban bulrush base threshold was estimated at - 4 C. There was a difference of 10 and 100 ADD between the predicted and observed peak biomass occurrence of water hyacinth and Cuban bulrush, respectively. Models suggest Cuban bulrush can survive lower temperatures than water hyacinth and can potentially invade states farther north than its current distribution in the United States. It is likely that annual winter ice formation on water bodies will be the major barrier to northern expansion Cuban bulrush.

Literature describing effective control measures for the floating-leaved plants American lotus (Nelumbo lutea Willd.), white waterlily (Nymphaea odorata Aiton), and watershield (Brasenia schreberi J.F. Gmel.) is minimal as these are usually considered as desirable species. However, floating-leaved plants can cause ecological, economic, and social problems when undergoing demographic expansions, usually following alterations of natural hydrologic cycles. Therefore, a mesocosm trial was conducted to determine the potential of foliar applications of seven aquatic herbicides to reduce abundance of the three target species at maximum and half-maximum label rates. Three of the herbicides (glyphosate, imazamox, and florpyrauxifen-benzyl) provided short- and long-term suppression (.75% reduction) of white waterlily and watershield leaf density or biomass. As a followup trial, field work was conducted using glyphosate, imazamox, and florpyrauxifen-benzyl to determine plant response to these herbicides in a natural setting. All herbicides resulted in long-term (52 wk after treatment) leaf density reduction of white waterlily (64 to 100% reduction) and watershield (46 to 75% reduction; except 2.83 kg ae ha 1 glyphosate) in field sites while the abundance of American lotus increased. Reduction of white waterlily and watershield may have reduced competition thereby favoring higher abundance of lotus. Regardless, long-term (52-wk) reduction of white waterlily and watershield suggest the potential for these herbicides as operational management tools for nuisance populations of these species. Future work should evaluate chemical techniques for control of American lotus, where both timing of leaf emergence and potential interactions with other plant species must be considered in the design of those studies.

Plant communities of aquatic ecosystems have outsized effects on the system structure and function. In Mississippi, aquatic plant communities are often poorly described, particularly in small lakes. In June 2024, the plant community of Mosquito Run at Matthews Brake National Wildlife Refuge was described in Leflore County, Mississippi, using a littoral zone point survey. Mosquito Run is a cypress- tupelo gum backswamp whose hydrology is dominated by fluvial processes within the Mississippi Alluvial Plain. This survey described a species rich aquatic plant community with substantial infestation of invasive species. The dominant submersed macrophyte, however, was Utricularia macrorhiza, a native aquatic plant which was previously presumed absent in Mississippi. This observation acts as the first record of this species in Leflore County, Mississippi and the greater Mississippi Delta. These findings assert the importance of continue floristic surveys, particularly of aquatic ecosystems, in Mississippi and the greater southeastern United States.

Phenological studies for Cuban bulrush [Oxycaryum cubense (Poepp. & Kunth) Lye] have been limited to the monocephalous form in Lake Columbus (Mississippi). Accordingly, there is little available information on potential phenological differences among O. cubense forms (monocephalous vs. polycephalous) and populations in other geographic locations in the United States. Therefore, seasonal patterns of biomass and starch allocation in O. cubense were quantified from two populations in Lake Columbus on the Tennessee-Tombigbee Waterway in Mississippi (monocephalous), two populations from Lake Martin in Louisiana (polycephalous), and two populations from Orange Lake in Florida (polycephalous). Monthly samples of O. cubense inflorescence, emergent, and submersed tissue were harvested from two plots per state from October 2021 to September 2022. During monthly data collection, air temperature and photoperiod were recorded. Starch allocation patterns were similar among all sites, with starch storage being less than 1.5% dry weight for all plant tissues. Biomass was greatest in Lake Columbus (monocephalous; 600.7 g dry weight [DW] m(-2)) followed by Lake Martin (polycephalous; 392.3 g DW m(-2)) and Orange Lake (polycephalous; 233.85 g DW m(-2)). Peak inflorescence biomass occurred in the winter for the Lake Martin and Orange Lake populations and in the summer for the Lake Columbus population. Inflorescence biomass in Lake Columbus had a positive relationship (r(2) = 0.53) with warmer air temperatures. Emergent and submersed biomass generally had negative relationships with both photoperiod and temperature (r(2) = 0.02 to 0.77) in all sites. Peak biomass was also negatively related to temperature and photoperiod. Results from this study indicate that there are differences in biomass allocation between the two growth forms of O. cubense and that growth can occur at temperatures below freezing. Low temperature tolerance may allow this species to expand its range farther north than previously suspected.